From isac.sehlstedt at psy.gu.se Mon Oct 2 13:03:17 2017 From: isac.sehlstedt at psy.gu.se (Isac Sehlstedt) Date: Mon, 2 Oct 2017 11:03:17 +0000 Subject: [FieldTrip] Follow up question: Computing pca variables (i.e. Latent, and coefficient variables) after ft_componentanalysis (Schoffelen, J.M. (Jan Mathijs)) Message-ID: Dear Jan, I am such a blind hen sometimes. Not the proudest moment of my life. Thank you for helping me find what I was looking for. Very Best, Isac -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 09:33:12 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 07:33:12 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> Message-ID: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 09:39:04 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 07:39:04 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization Message-ID: <6d86574c600e4bc4a114b4d7ea788d18@neuro.gu.se> Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz -------------- next part -------------- An HTML attachment was scrubbed... URL: From sarang at cfin.au.dk Wed Oct 4 09:47:19 2017 From: sarang at cfin.au.dk (Sarang S. Dalal) Date: Wed, 4 Oct 2017 07:47:19 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> Message-ID: <1507103238.2609.8.camel@cfin.au.dk> Hi Bushra, Are you using the common filter approach to use the same spatial filter for your contrasted conditions? http://www.fieldtriptoolbox.org/example/common_filters_in_beamforming (Although that example is focused on DICS and PCC, a similar approach can also be used for LCMV.) This should reduce the likelihood of the 'middle of the head' artifact. If the filters are computed independently, however, then the noise suppression characteristics of each filter can be different, so the contrast would pull out differences in each filter's noise performance. If you are already using this approach, then you can try weight normalization, which is one strategy for removing the noise bias (and usually prevents the 'center of the head' artifact even when no contrast is done). You can run this by adding to your cfg structure: cfg.lcmv.weightnorm = 'nai' - or - cfg.lcmv.weightnorm = 'unitnoisegain'; Each of those methods should give you the same result with a contrast. The NAI is simply scaled against the theoretical noise power (i.e., the noise level = 1), so is often preferred when no contrast is done. Cheers, Sarang On Wed, 2017-10-04 at 07:33 +0000, Bushra Riaz Syeda wrote: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don’t see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 10:19:09 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 08:19:09 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: <1507103238.2609.8.camel@cfin.au.dk> References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> , <1507103238.2609.8.camel@cfin.au.dk> Message-ID: Thank you for you reply I am already using common filter for both conditions and I have also tried weight normalisation with 'unitnoisegain', but it didn't change results. My protocol is bit unusual in a way that electric stimulation to the finger is time locked with 200 ms after the R wave of ECG. As such, the ECG artifact is prominent at -200 ms in my data and I am using ICA to remove it as shown in figure (https://chalmersuniversity.box.com/s/deuzucqovgcfienybk7zisc3r3wjusvu) For lcmv I am using -0.71 to -0.21 sec of prestimulus interval window ( relatively clean baseline without ECG artifact) and 0.01 to 0.51 sec of post stimulus interval window. Could my protocol be the reason of 'center of head artifact', I didnt think it would be the case? Any suggestions to solve this would be highly appreciated. Thank you Kind regards Bushra Riaz ________________________________ From: fieldtrip-bounces at science.ru.nl on behalf of Sarang S. Dalal Sent: Wednesday, October 4, 2017 9:47:19 AM To: fieldtrip at science.ru.nl Subject: Re: [FieldTrip] Artifact in LCMV beamformer source localization Hi Bushra, Are you using the common filter approach to use the same spatial filter for your contrasted conditions? http://www.fieldtriptoolbox.org/example/common_filters_in_beamforming (Although that example is focused on DICS and PCC, a similar approach can also be used for LCMV.) This should reduce the likelihood of the 'middle of the head' artifact. If the filters are computed independently, however, then the noise suppression characteristics of each filter can be different, so the contrast would pull out differences in each filter's noise performance. If you are already using this approach, then you can try weight normalization, which is one strategy for removing the noise bias (and usually prevents the 'center of the head' artifact even when no contrast is done). You can run this by adding to your cfg structure: cfg.lcmv.weightnorm = 'nai' - or - cfg.lcmv.weightnorm = 'unitnoisegain'; Each of those methods should give you the same result with a contrast. The NAI is simply scaled against the theoretical noise power (i.e., the noise level = 1), so is often preferred when no contrast is done. Cheers, Sarang On Wed, 2017-10-04 at 07:33 +0000, Bushra Riaz Syeda wrote: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From na.so.ir at gmail.com Thu Oct 5 12:44:35 2017 From: na.so.ir at gmail.com (Narjes Soltani) Date: Thu, 5 Oct 2017 12:44:35 +0200 Subject: [FieldTrip] Median Filter Message-ID: Dear Fieldtripers, I have a question about the median filter. When I set this filter to 'no' (which is the default value) in ft_preprocessing configuration file and then run the ICA on the output, multiple components are not converged. However if I set the median filter to 'yes', this problem vanishes. Is there any idea why setting the median filter to 'yes' which means keeping the jump artifacts, helps in ICA convergence? By the way, I sent the epoched data to ft_preprocessing which may also be important about how ICA works. Very Best, Narjes -------------- next part -------------- An HTML attachment was scrubbed... URL: From chrichat at hotmail.com Fri Oct 6 15:42:02 2017 From: chrichat at hotmail.com (Christos Chatzichristos) Date: Fri, 6 Oct 2017 13:42:02 +0000 Subject: [FieldTrip] Simulation of sources based on fMRI spatial map. Message-ID: Dear all I am new at the community of field trip (and EEG ) generally. I am a PHD student and I am working on tensor representation for fMRI Blind Source Separation. Currently I am working on a parallel project for fusion of EEG and fMRI (with tensors again). Since there are new methods before going to real data I need some realistic simulated to test my methods on. I went through the tutorials about `Creating a sourcemodel for source-reconstruction of MEG or EEG data' and also the example 'Compute forward simulated data and apply a dipole fit' but still I was wondering if there is a way to create the leadfield matrices for sources simulated for fMRI or even more simply, from single slice sources like the one attached (which has been simulated with SimTB for a spherical brain). I was thinking to perform a simulation similar to this used in [X. Lei et al. 'A parallel framework for simultaneous EEG/fMRI analysis: Methodology and simulation], where they build a concentric three-sphere model around a spherical brain slice similar to the one I attach. Thanks a lot for any possible help. Kind regards Christos Chatzichristos -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Screenshot from 2017-10-06 14-37-38.png Type: image/png Size: 17053 bytes Desc: Screenshot from 2017-10-06 14-37-38.png URL: From aarjona at us.es Fri Oct 6 12:09:48 2017 From: aarjona at us.es (Antonio Arjona Valladares) Date: Fri, 06 Oct 2017 12:09:48 +0200 Subject: [FieldTrip] Fwd: the field cfg.neighbours is required In-Reply-To: References: Message-ID: Hello everyone! This is my query: Im trying to make statistical analyses of time-frequency with two groups of subjects. I have created the file .study with both groups, precompute de channels, and then I would like to apply the 'montecarlo/permutation statistics' with 'cluster correction'. The problem comes when I select all the electrodes and click on 'Plot ERSP(s)'. After a few seconds, the program gives me this answer (image 3). I have been reading tutorials, but nothings appears about this issue. Thank you so much --- Antonio Arjona Valladares Research Technician (BS) Human Psychobiology Lab, Experimental Psychology Department, University of Seville, C/Camilo José Cela s/n, 41018-Sevilla, Spain Tel: +34 954 55 78 00 - 954 55 69 41 -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image 3.png Type: image/png Size: 1067786 bytes Desc: not available URL: From abela.eugenio at gmail.com Fri Oct 6 17:30:50 2017 From: abela.eugenio at gmail.com (Eugenio Abela) Date: Fri, 6 Oct 2017 16:30:50 +0100 Subject: [FieldTrip] Fwd: the field cfg.neighbours is required In-Reply-To: References: Message-ID: <9A6240C6-2D77-4A26-BB76-3DDFD61CD417@gmail.com> Hi Antonio this looks like a question that EEGLAB might also be able to answer (https://sccn.ucsd.edu/wiki/EEGLAB_mailing_lists )? In any case, see here for what cfg.neighbours means /does: http://www.fieldtriptoolbox.org/tutorial/cluster_permutation_timelock http://www.fieldtriptoolbox.org/faq/how_can_i_define_neighbouring_sensors I’m sorry, but I don’t know how you specify the neighbour structure in EEGLAB... Good luck Eugenio On 6 Oct 2017, at 11:09, Antonio Arjona Valladares wrote: Hello everyone! This is my query: Im trying to make statistical analyses of time-frequency with two groups of subjects. I have created the file .study with both groups, precompute de channels, and then I would like to apply the 'montecarlo/permutation statistics' with 'cluster correction'. The problem comes when I select all the electrodes and click on 'Plot ERSP(s)'. After a few seconds, the program gives me this answer (image 3). I have been reading tutorials, but nothings appears about this issue. Thank you so much --- Antonio Arjona Valladares Research Technician (BS) Human Psychobiology Lab, Experimental Psychology Department, University of Seville, C/Camilo José Cela s/n, 41018-Sevilla, Spain Tel: +34 954 55 78 00 - 954 55 69 41 _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image 3.png Type: image/png Size: 231844 bytes Desc: not available URL: From cmhan616 at utexas.edu Tue Oct 10 21:54:57 2017 From: cmhan616 at utexas.edu (Chungmin Han) Date: Tue, 10 Oct 2017 14:54:57 -0500 Subject: [FieldTrip] Question on TFR analysis Message-ID: Dear all, I'm Chungmin, graduate student in UT Austin, and have a question TFR using fieldtrip, my experiment has 2 triggers, first one is Trial onset second trigger is the beginning of task with fixed time when subject trigger to begin task. Thus there's varying time between 1st and 2nd trigger, I need to analyze signal with fixed time window after 2nd trigger but need to take baseline right before 1st trigger because there's movement between 1st and 2nd, is there a way I can set up baseline window in matrix form?? seems TFRanalysis allows single baseline time window settings not for varying time window, I've also looked at redefine_trials, however, in this case, I'll have to put two inputs to calculate TFR, one as baselines only and the other task only. .....if either of the methods is not possible, can I set up single baseline (e.g. resting state data)??? Any advice is appreciated Thank you, Best Chungmin -------------- next part -------------- An HTML attachment was scrubbed... URL: From M.vanEs at donders.ru.nl Wed Oct 11 13:58:14 2017 From: M.vanEs at donders.ru.nl (Es, M.W.J. van (Mats)) Date: Wed, 11 Oct 2017 11:58:14 +0000 Subject: [FieldTrip] Question on TFR analysis (Chungmin Han) Message-ID: <3FC79061C73BEF44A3BEDA5DFC0ADBDF86529C4F@EXPRD99.hosting.ru.nl> Hi Chungmin, You are right, it is not possible to make your contrast in a single step. It is not possible to do a baseline correction in ft_freqbaseline (also holds for ft_singleplotTFR etc.) when the baseline is time locked to a different event than the 'active' condition. In order to get the contrast you should get the TFR for both 'active' and 'baseline' condition, timelocked to the different events (by using ft_redefinetrial). You can then use ft_math to calculate the difference. Good luck, Mats van Es -----Original Message----- From: fieldtrip-bounces at science.ru.nl [mailto:fieldtrip-bounces at science.ru.nl] On Behalf Of fieldtrip-request at science.ru.nl Sent: woensdag 11 oktober 2017 12:00 To: fieldtrip at science.ru.nl Subject: fieldtrip Digest, Vol 83, Issue 6 Send fieldtrip mailing list submissions to fieldtrip at science.ru.nl To subscribe or unsubscribe via the World Wide Web, visit https://mailman.science.ru.nl/mailman/listinfo/fieldtrip or, via email, send a message with subject or body 'help' to fieldtrip-request at science.ru.nl You can reach the person managing the list at fieldtrip-owner at science.ru.nl When replying, please edit your Subject line so it is more specific than "Re: Contents of fieldtrip digest..." Today's Topics: 1. Question on TFR analysis (Chungmin Han) ---------------------------------------------------------------------- Message: 1 Date: Tue, 10 Oct 2017 14:54:57 -0500 From: Chungmin Han To: fieldtrip at science.ru.nl Subject: [FieldTrip] Question on TFR analysis Message-ID: Content-Type: text/plain; charset="utf-8" Dear all, I'm Chungmin, graduate student in UT Austin, and have a question TFR using fieldtrip, my experiment has 2 triggers, first one is Trial onset second trigger is the beginning of task with fixed time when subject trigger to begin task. Thus there's varying time between 1st and 2nd trigger, I need to analyze signal with fixed time window after 2nd trigger but need to take baseline right before 1st trigger because there's movement between 1st and 2nd, is there a way I can set up baseline window in matrix form?? seems TFRanalysis allows single baseline time window settings not for varying time window, I've also looked at redefine_trials, however, in this case, I'll have to put two inputs to calculate TFR, one as baselines only and the other task only. .....if either of the methods is not possible, can I set up single baseline (e.g. resting state data)??? Any advice is appreciated Thank you, Best Chungmin -------------- next part -------------- An HTML attachment was scrubbed... URL: ------------------------------ _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip End of fieldtrip Digest, Vol 83, Issue 6 **************************************** From luca.turella at unitn.it Wed Oct 11 17:06:13 2017 From: luca.turella at unitn.it (Luca Turella) Date: Wed, 11 Oct 2017 17:06:13 +0200 Subject: [FieldTrip] Post-doc in Neural dynamics of Action understanding @ CIMeC University of Trento, Italy Message-ID: A postdoctoral position will be available soon at the Center for Mind/Brain Sciences (CIMeC, http://www.cimec.unitn.it/en) at the University of Trento (Italy). The topic of investigation will cover the neural dynamics underlying action understanding adopting behavioural studies and MEG. The requirements for the candidates are the following. Candidates must have a Ph.D. degree in a field related to Cognitive Neuroscience or related areas. Previous experience in behavioural/kinematic studies and/or EEG/MEG data analysis is required. The ideal candidate should have good programming skills in Matlab, preferably related to EEG/MEG data analysis (Fieldtrip). Knowledge of kinematic data analysis is also a plus. Knowledge of Italian language is not required. The salary will be proportional to the level of experience and the starting date of the appointment is negotiable, but within the next 6 months. Applications will be considered until the position is filled. The contract will have a duration of 2 years. Applications should be sent to luca.turella at unitn.it, including complete CV, statement of research interests, and contact details of two referees. Potential candidates are also encouraged to send me directly informal inquires *luca.turella at unitn.it *. CIMeC offers an international and vibrant research setting with access to state-of-the-art neuroimaging methodologies, including a research-only MR scanner, MEG, EEG and TMS, as well as behavioural, eye tracking and motion tracking laboratories. English is the official language of the CIMeC, where a large proportion of the faculty, post-docs and students come from a wide range of countries outside of Italy. The University of Trento consistently ranks as a top Italian university in both national Research Assessment Evaluations (RAE) and University Surveys. In the latest RAE, the University of Trento as a whole ranks 2nd among medium-sized universities. -- Luca Turella, PhD Assistant Professor CIMeC - Center for Mind/Brain Sciences University of Trento Mattarello (TN), Via Delle Regole 101 Tel.+39 0461-28 3098 http://www.unitn.it/cimec Legal Disclaimer This electronic message contains information that is confidential. The information is intended for the use of the addressee only. If you are not the addressee we would appreciate your notification in this respect. Please note that any disclosure, copy, distribution or use of the contents of this message is prohibited and may be unlawful. Avvertenza legale Questo messaggio Email contiene informazioni confidenziali riservate ai soli destinatari. Qualora veniate in possesso di tali informazioni senza essere definito come destinatario vi reghiamo di leggere le seguenti note. Ogni apertura, copia, distribuzione del contenuto del messaggio e dei suoi allegati è proibito e potrebbe violare le presenti leggi. -------------- next part -------------- An HTML attachment was scrubbed... URL: From m.stoica at uke.de Wed Oct 11 21:54:41 2017 From: m.stoica at uke.de (Mircea Stoica) Date: Wed, 11 Oct 2017 21:54:41 +0200 Subject: [FieldTrip] Sharp transition in multitaper coherence when moving from one taper to two Message-ID: Dear fieldtrippers, I'm interested in coherence analysis with mtmconvol and multitapers. My frequencies of interest are spaced logarithmically (steps of 1/16 octaves) and I adjust the time window and number of tapers to yield a constant frequency smoothing of 0.75 octaves (half that for cfg.tapsmofrq of course). This results in an ugly step-like transition of coherence when moving from one taper to two, as you can see in the following image. https://photos.app.goo.gl/5uWy05RzZxaA9gci1 This sharp transition is not visible in the power output. I calculated the actual frequency smoothing as fw = (K + 1) ./ cfg.t_ftimwin / 2 and in a semilogy plot it gives this: https://photos.app.goo.gl/xQsrY1jr8L0qLLDm2 And the number of tapers for each frequency: https://photos.app.goo.gl/LqaeUVW6ymBWEKEx2 There are no discontinuities in the frequency smoothing. Any ideas what could be the cause? Or is there perhaps nothing to be done about it? Best, Mircea -- _____________________________________________________________________ Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg | www.uke.de Vorstandsmitglieder: Prof. Dr. Burkhard Göke (Vorsitzender), Prof. Dr. Dr. Uwe Koch-Gromus, Joachim Prölß, Martina Saurin (komm.) _____________________________________________________________________ SAVE PAPER - THINK BEFORE PRINTING -------------- next part -------------- An HTML attachment was scrubbed... URL: From stephen.whitmarsh at gmail.com Fri Oct 13 17:29:35 2017 From: stephen.whitmarsh at gmail.com (Stephen Whitmarsh) Date: Fri, 13 Oct 2017 17:29:35 +0200 Subject: [FieldTrip] NAI on trial-by-trail power estimates Message-ID: Dear all, I am trying to optimize my beamformer power estimates, which are not bad, but I want to see if I can improve them. I do not seem to have a depth-bias, and am testing within-subject between-condition contrasts, for which one does not need to correct for noise differences. However, my analysis does involve trial-by-trial analysis, and might be susceptible to noise differences over time. I was wondering whether it would make sense to have trial-by-trial corrections using the Neural Activation Index (NAI), as described in the beamformer tutorial, i.e. by dividing my single-trial power estimates by single-trial noise estimates. Has anyone tried using either the NAI on a trial-by-trial basis? Secondly, does this even make sense to you? Cheers, Stephen -------------- next part -------------- An HTML attachment was scrubbed... URL: From yuxxx955 at umn.edu Mon Oct 16 23:26:30 2017 From: yuxxx955 at umn.edu (Kai Yu) Date: Mon, 16 Oct 2017 16:26:30 -0500 Subject: [FieldTrip] How to understand the cross-correlogram and its feature comparison against the one obtained by shuffling the trials Message-ID: Hi, I have been using your FieldTrip Intracranial Spike Analysis package for a while. However, I am still confused by the statement listed on your website. "Cross-correlations between neurons can either arise because of common, time-locked fluctuations in the firing rate (Brody et al., 1999). These correlations are invariant to a change in the order of trials. The shuffling of trials in ft_spike_xcorr always pertains to two subsequent trials, in order to avoid an influence of slow changes in the firing rate across trials. We refer to this cross-correlogram that is obtained under a permutation of subsequent trials as the 'shift-predictor' cross-correlogram. If the observed features of the cross-correlogram that are not present in the shift-predictor cross-correlogram, then this indicates that they arise because of induced synchronous activity." For the last sentence, would you please explained more about the "induced synchronous activity"? In my own understanding, this is the activity after the task or external stimulation. But if I want to know the synchronous or de-synchronous activity during the stimulation, how can I get the evidence from these cross-correlations and also the jpsth figures? Thank you so much! Best, -Kai -- Kai Yu Biomedical Functional Imaging and Neuroengineering Laboratory Department of Biomedical Engineering University of Minnesota - Twin Cities -------------- next part -------------- An HTML attachment was scrubbed... URL: From acskwara at ucdavis.edu Wed Oct 18 23:05:45 2017 From: acskwara at ucdavis.edu (Alea Skwara) Date: Wed, 18 Oct 2017 14:05:45 -0700 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) Message-ID: Hi All, I have been conducting a permutation based cluster analysis using the method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to examine longitudinal changes in spectral power. Everything runs just fine and it identifies clusters successfully. However, when I examine the results of the analysis, some of the stats strike me as strange. Electrodes that have the highest F values (I am comparing across 3 timepoints) do not test as significant, while electrodes with lower F values do. At first I thought this was because I had a single "hot" electrode that was not part of a larger cluster, but looking more closely at the stats, it appears that this is not the case. I have tried double-checking my neighbor definition method and electrode locations to assure that it is not an issue with neighbors being incorrectly registered. I have also tried adjusting the minimum cluster size using the minnbchan option to ensure it was not an issue of a cluster not reaching the minimum extent. Neither of these options seem to have any impact on the result. At this point I'm at a loss. Is there something I'm misunderstanding about the algorithm? Code below and plot below, channel-by-channel stats attached. Thanks! Alea One funky thing to note about this code is that I am inputting average values into the analysis (averaged across time as participants have slightly differing lengths of data, and within bands as I am running this analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band power", not 7 Hz) %% THREE CONDITIONS %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) cfg = []; cfg.channel = {'all'}; cfg.latency = 'all'; cfg.frequency = [7 7]; %frequency band indicator cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing to average cfg.method = 'montecarlo'; cfg.statistic = 'depsamplesFmultivariate'; %three time points cfg.correctm = 'cluster'; cfg.clusteralpha = 0.1; %change to .1 because we're computing over averaged bands and time (losing power) cfg.clusterstatistic = 'maxsum'; cfg.minnbchan = 2; cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 otherwise cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account for 500 tests and correction by Bonferroni method. %specify which electrodes can form clusters cfg_neighb.method = 'distance'; cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); %cfg.neighbours = cfgManish.neighbours; %cfg = getNeighbours(cfg); % using fixed neighbours cfg.layout = layout'; % Set up the design matrixfor 3 conditions subj = 27; %will differ based on condition design = zeros(2,3*subj); for isub = 1:1:subj design(1,isub) = isub; design(1,subj+isub) = isub; design(1,2*subj+isub) = isub; end design(2,1:subj) = 1; design(2,subj+1:2*subj) = 2; design(2,2*subj+1:3*subj) = 3; cfg.design = design; cfg.uvar = 1; cfg.ivar = 2; % Run the analysis [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, r1t3r_struct); %Plot the results cfg1 = []; cfg1.alpha = 0.1; cfg1.parameter = 'stat'; %cfg1.zlim = [0 18.1]; cfg1.layout = layout'; cfg1.marker = 'o'; cfg1.subplotsize = [1 1]; cfg1.saveaspng = 'r1r_betacluster'; ft_clusterplot(cfg1, r1r_stat_beta); ***Note non-significant hot spot over left parietal*** [image: Inline image 2] Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: r1r_betacluster_fig1.png Type: image/png Size: 49573 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: clusterstats_beta.xlsx Type: application/vnd.openxmlformats-officedocument.spreadsheetml.sheet Size: 11196 bytes Desc: not available URL: From bioeng.yoosofzadeh at gmail.com Thu Oct 19 05:34:09 2017 From: bioeng.yoosofzadeh at gmail.com (Vahab Yousofzadeh) Date: Wed, 18 Oct 2017 23:34:09 -0400 Subject: [FieldTrip] Postdoctoral position at Ulster on Dementia and neuroimaging Message-ID: Hi everyone, There is a postdoc position at Ulster Univerity with Dr. KongFatt Wong-Lin and Prof. Tony Bjourson on dementia and neuroimaging. Check it out: https://goo.gl/4YexD5 Cheers, Vahab From jan.schoffelen at donders.ru.nl Thu Oct 19 08:19:35 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Thu, 19 Oct 2017 06:19:35 +0000 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) In-Reply-To: References: Message-ID: Hi Alea, I think that it all looks OK. I guess that the P5-PO1-O1 cluster is spatially disjoint from the bulk of the suprathreshold data points. Don’t be mislead by the topographical image, where both the contour lines, and the 2D placement of the electrodes might reflect a picture that is different from what you would expect. Note, that providing cirange as extra information is not the most informative. Better would have been to show posclusterslabelmat. This data field provides for each sample, whether and to which ‘cluster’ it belongs. Best wishes, Jan-Mathijs J.M.Schoffelen, MD PhD Senior Researcher, VIDI-fellow - PI, language in interaction Telephone: +31-24-3614793 Physical location: room 00.028 Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands On 18 Oct 2017, at 23:05, Alea Skwara > wrote: Hi All, I have been conducting a permutation based cluster analysis using the method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to examine longitudinal changes in spectral power. Everything runs just fine and it identifies clusters successfully. However, when I examine the results of the analysis, some of the stats strike me as strange. Electrodes that have the highest F values (I am comparing across 3 timepoints) do not test as significant, while electrodes with lower F values do. At first I thought this was because I had a single "hot" electrode that was not part of a larger cluster, but looking more closely at the stats, it appears that this is not the case. I have tried double-checking my neighbor definition method and electrode locations to assure that it is not an issue with neighbors being incorrectly registered. I have also tried adjusting the minimum cluster size using the minnbchan option to ensure it was not an issue of a cluster not reaching the minimum extent. Neither of these options seem to have any impact on the result. At this point I'm at a loss. Is there something I'm misunderstanding about the algorithm? Code below and plot below, channel-by-channel stats attached. Thanks! Alea One funky thing to note about this code is that I am inputting average values into the analysis (averaged across time as participants have slightly differing lengths of data, and within bands as I am running this analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band power", not 7 Hz) %% THREE CONDITIONS %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) cfg = []; cfg.channel = {'all'}; cfg.latency = 'all'; cfg.frequency = [7 7]; %frequency band indicator cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing to average cfg.method = 'montecarlo'; cfg.statistic = 'depsamplesFmultivariate'; %three time points cfg.correctm = 'cluster'; cfg.clusteralpha = 0.1; %change to .1 because we're computing over averaged bands and time (losing power) cfg.clusterstatistic = 'maxsum'; cfg.minnbchan = 2; cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 otherwise cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account for 500 tests and correction by Bonferroni method. %specify which electrodes can form clusters cfg_neighb.method = 'distance'; cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); %cfg.neighbours = cfgManish.neighbours; %cfg = getNeighbours(cfg); % using fixed neighbours cfg.layout = layout'; % Set up the design matrixfor 3 conditions subj = 27; %will differ based on condition design = zeros(2,3*subj); for isub = 1:1:subj design(1,isub) = isub; design(1,subj+isub) = isub; design(1,2*subj+isub) = isub; end design(2,1:subj) = 1; design(2,subj+1:2*subj) = 2; design(2,2*subj+1:3*subj) = 3; cfg.design = design; cfg.uvar = 1; cfg.ivar = 2; % Run the analysis [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, r1t3r_struct); %Plot the results cfg1 = []; cfg1.alpha = 0.1; cfg1.parameter = 'stat'; %cfg1.zlim = [0 18.1]; cfg1.layout = layout'; cfg1.marker = 'o'; cfg1.subplotsize = [1 1]; cfg1.saveaspng = 'r1r_betacluster'; ft_clusterplot(cfg1, r1r_stat_beta); ***Note non-significant hot spot over left parietal*** Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From maximilien.chaumon at gmail.com Thu Oct 19 11:18:51 2017 From: maximilien.chaumon at gmail.com (Maximilien Chaumon) Date: Thu, 19 Oct 2017 09:18:51 +0000 Subject: [FieldTrip] memory leak in ft_selectdata Message-ID: Dear all, I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion cfg = []; cfg.parameter = 'logmpowspctrm'; cfg.comment = 'no'; cfg.xlim = [20 21]; cfg.zlim = 'maxmin'; cfg.layout = 'neuromag306all.lay'; cfg.colorbar = 'no'; big_spec_grad.dimord = 'chan_freq'; ft_topoplotER(cfg,big_spec_grad); This used to work fine until a last update this week from a version I suspect dated back to 03/17. Now today, the ft_topoplotER line eats up a lot of memory the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. x = x(:,selindx,:,:,:,:); The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... Any help is much appreciated. Thanks! Max -------------- next part -------------- An HTML attachment was scrubbed... URL: From adonay.s.nunes at gmail.com Thu Oct 19 23:41:53 2017 From: adonay.s.nunes at gmail.com (A Nunes) Date: Thu, 19 Oct 2017 14:41:53 -0700 Subject: [FieldTrip] SAM gareth method without depth biased weights Message-ID: Hi community, I have an issue with Fieltrip SAM beamformer 'gareth' method. As a sanity check I look that the BF weight distributions are biased towards the center. I check this distribution with the leadfields and with the BF weights from LCMV and SAMs methods. Gareth method does not have this distribution, the weight sizes are scattered across the brain and the norm of the weights can have up to 11 orders of magnitude of difference. First I make sure that the 1/Leadfield are also biased towards the center: - for every LF I take the norm of every column and 1/mean of the norms is the value. - the code is: LF_norm = []; for s= 1:size(LF, 1) n1 = norm(squeeze(LF(s, 1, :))); n2 = norm(squeeze(LF(s, 2, :))); n3 = norm(squeeze(LF(s, 3, :))); LF_norm(s) =1/mean([n1, n2, n3]); end When interpolated to the mri and plotted looks fine. Then I run SAM robert method: cfg = []; cfg.method = 'sam'; cfg.grid = lf; cfg.vol = headmodel; cfg.grad = data.grad; cfg.fixedori = 'robert'; cfg.keepfilter = 'yes'; cfg.keepori = 'yes'; cfg.projectnoise = 'yes'; cfg.senstype = 'MEG'; SAM1_filter = ft_sourceanalysis(cfg, timelock); and take the norm for every BF weight: SAM1_filter_norm= cellfun(@(X) norm(X),SAM1_filter.avg.filter(SAM1_filter.inside)); and then I interpolate it to the mri for visualization. The results look fine, weights biased towards the center and the min and max norm weight are two orders of magnitude difference. The mean of the absolute of the weights norm is in the order of 10e9 and the smallest and highest weight norm are to the order of e8 and e10 If I do the same as the last step but changing to: cfg.fixedori = 'gareth'; Then the weights are not biased towards the center. The mean of the absolute of the weights norm is in the order of 10e16 and the smallest and highest weight norm are to the order e8 and e19, respectively. Thus, gareth method does not have a center biased weight distribution and the max and miminum norm weights have 11 orders of magnitude of difference, meaning that these orders are reflected in the amplitude of the reconstructed time series. To make it handy the difference between gareth and robert methods are: L is the leadfield for a given source. 'Gareth' Y1 = L' * inv_cov * L; Y2 = L' * (inv_cov * inv_cov) * L; [U,S] = eig(Y2,Y1); 'Robert' [U,S] = svd(real(pinv(L' * inv_cov * L))); Then both methods compute 2 of the orientations and choose the salient one: ori1 = U(:,1); ori1 = ori1/norm(ori1); ori2 = U(:,2); ori2 = ori2/norm(ori2); L1 = L * ori1; L2 = L * ori2; if (norm(L1)/norm(L2)) < 1e-6 opt_vox_or = ori2; else opt_vox_or = ori1; end The difference comes from the way orientations are computed, after the weights calculation is the same for both. I tried with different units, i.e. meters, cm, mm but this abnormalities is still present. Does anybody know what can be the cause? Thanks Adonay -------------- next part -------------- An HTML attachment was scrubbed... URL: From jan.schoffelen at donders.ru.nl Fri Oct 20 09:07:10 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Fri, 20 Oct 2017 07:07:10 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: Hi Max, >From your (otherwise clear) description, I am not sure whether your issue is due to the actual data selection itself. Rather, my first thought would be that something goes amiss with recursive piling up of large amounts of irrelevant (for the plotting at least) metadata, that gets assembled while iterating through subsequent interactive plotting steps. It’s kind of a long story, which I am not going to type down here, but to check whether I am thinking into the right direction, could you see whether your problem persists if you call ft_topoplotER in the following way? ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); Thanks, Jan-Mathijs > On 19 Oct 2017, at 11:18, Maximilien Chaumon wrote: > > Dear all, > > I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion > cfg = []; > cfg.parameter = 'logmpowspctrm'; > cfg.comment = 'no'; > cfg.xlim = [20 21]; > cfg.zlim = 'maxmin'; > cfg.layout = 'neuromag306all.lay'; > cfg.colorbar = 'no'; > big_spec_grad.dimord = 'chan_freq'; > ft_topoplotER(cfg,big_spec_grad); > > This used to work fine until a last update this week from a version I suspect dated back to 03/17. > Now today, the ft_topoplotER line eats up a lot of memory > the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB > the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB > the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB > > This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. > > I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. > x = x(:,selindx,:,:,:,:); > The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). > Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... > > Any help is much appreciated. > Thanks! > Max > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip From mibanks at wisc.edu Fri Oct 20 16:16:54 2017 From: mibanks at wisc.edu (MATTHEW I BANKS) Date: Fri, 20 Oct 2017 14:16:54 +0000 Subject: [FieldTrip] Variance estimates for connectivity measures Message-ID: Greetings. Is it possible to use jackknife to get variance estimates for Granger, partial directed coherence and directed transfer function? I use it for wPLI, but cannot figure out how to do it for these other measures. -Matt Banks ____________________________ Matthew I. Banks, Ph.D. Associate Professor Department of Anesthesiology University of Wisconsin 1300 University Avenue, Room 4605 Madison, WI 53706 office tel. (608)261-1143 lab tel. (608)263-6662 fax (608)263-2592 http://anesthesia.wisc.edu/index.php/Banks_Laboratory http://ntp.neuroscience.wisc.edu/banks.htm -------------- next part -------------- An HTML attachment was scrubbed... URL: From maximilien.chaumon at gmail.com Fri Oct 20 16:30:19 2017 From: maximilien.chaumon at gmail.com (Maximilien Chaumon) Date: Fri, 20 Oct 2017 14:30:19 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: Perfect! that fixed it, many thanks! Max Le ven. 20 oct. 2017 à 09:35, Schoffelen, J.M. (Jan Mathijs) < jan.schoffelen at donders.ru.nl> a écrit : > Hi Max, > > From your (otherwise clear) description, I am not sure whether your issue > is due to the actual data selection itself. Rather, my first thought would > be that something goes amiss with recursive piling up of large amounts of > irrelevant (for the plotting at least) metadata, that gets assembled while > iterating through subsequent interactive plotting steps. > It’s kind of a long story, which I am not going to type down here, but to > check whether I am thinking into the right direction, could you see whether > your problem persists if you call ft_topoplotER in the following way? > > > ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); > > Thanks, > > Jan-Mathijs > > > > > On 19 Oct 2017, at 11:18, Maximilien Chaumon < > maximilien.chaumon at gmail.com> wrote: > > > > Dear all, > > > > I'm doing topoplots of rather large spectrum data focused on a tiny > frequency portion > > cfg = []; > > cfg.parameter = 'logmpowspctrm'; > > cfg.comment = 'no'; > > cfg.xlim = [20 21]; > > cfg.zlim = 'maxmin'; > > cfg.layout = 'neuromag306all.lay'; > > cfg.colorbar = 'no'; > > big_spec_grad.dimord = 'chan_freq'; > > ft_topoplotER(cfg,big_spec_grad); > > > > This used to work fine until a last update this week from a version I > suspect dated back to 03/17. > > Now today, the ft_topoplotER line eats up a lot of memory > > the call to "ft_prepare_layout" took 0 seconds and required the > additional allocation of an estimated 0 MB > > the call to "ft_selectdata" took 0 seconds and required the additional > allocation of an estimated 1283 MB > > the call to "ft_topoplotER" took 1 seconds and required the additional > allocation of an estimated 1283 MB > > > > This gets more dramatic the second time this part of the script runs > (with 2500MB added for reach topoplot) and then crashes on the next > iteration with all of 30GB RAM eaten up by this. > > > > I traced down the memory leak to ft_selectdata line 1281 in a > subfunction called cellmatselect. > > x = x(:,selindx,:,:,:,:); > > The memory usage jumps every time this dimension gets selected. (I admit > I do have a lot of frequencies, over 13000, but I need them for my > application). > > Memory comes back to a normal level when the figure is closed, but as I > said, execution with another dataset eats up double the amount of memory > and I don't know how to get out of this... > > > > Any help is much appreciated. > > Thanks! > > Max > > _______________________________________________ > > fieldtrip mailing list > > fieldtrip at donders.ru.nl > > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From jan.schoffelen at donders.ru.nl Fri Oct 20 16:40:34 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Fri, 20 Oct 2017 14:40:34 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: <68BED22E-0754-4991-B4D5-4F9EAE8536CE@donders.ru.nl> You’re welcome. Yet, I wouldn’t say it’s fixed, but at least you have found a way around your problem. Best wishes, Jan-Mathijs On 20 Oct 2017, at 16:30, Maximilien Chaumon > wrote: Perfect! that fixed it, many thanks! Max Le ven. 20 oct. 2017 à 09:35, Schoffelen, J.M. (Jan Mathijs) > a écrit : Hi Max, From your (otherwise clear) description, I am not sure whether your issue is due to the actual data selection itself. Rather, my first thought would be that something goes amiss with recursive piling up of large amounts of irrelevant (for the plotting at least) metadata, that gets assembled while iterating through subsequent interactive plotting steps. It’s kind of a long story, which I am not going to type down here, but to check whether I am thinking into the right direction, could you see whether your problem persists if you call ft_topoplotER in the following way? ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); Thanks, Jan-Mathijs > On 19 Oct 2017, at 11:18, Maximilien Chaumon > wrote: > > Dear all, > > I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion > cfg = []; > cfg.parameter = 'logmpowspctrm'; > cfg.comment = 'no'; > cfg.xlim = [20 21]; > cfg.zlim = 'maxmin'; > cfg.layout = 'neuromag306all.lay'; > cfg.colorbar = 'no'; > big_spec_grad.dimord = 'chan_freq'; > ft_topoplotER(cfg,big_spec_grad); > > This used to work fine until a last update this week from a version I suspect dated back to 03/17. > Now today, the ft_topoplotER line eats up a lot of memory > the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB > the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB > the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB > > This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. > > I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. > x = x(:,selindx,:,:,:,:); > The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). > Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... > > Any help is much appreciated. > Thanks! > Max > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From acskwara at ucdavis.edu Fri Oct 20 20:59:49 2017 From: acskwara at ucdavis.edu (Alea Skwara) Date: Fri, 20 Oct 2017 11:59:49 -0700 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) In-Reply-To: References: Message-ID: Hi J.M., Thanks so much for your reply. The tip to not trust the topographic maps too much helped in my sleuthing. I realized that adjusting the minnbchan parameter actually *does* result in the topographic changes I thought I should see if I was understanding the algorithm correctly (and that I just wasn't seeing these changes in the map that I was using as a quick-check, but that they were present in the stats themselves). Problem solved! Thank you! Alea On Wed, Oct 18, 2017 at 11:19 PM, Schoffelen, J.M. (Jan Mathijs) < jan.schoffelen at donders.ru.nl> wrote: > Hi Alea, > > I think that it all looks OK. I guess that the P5-PO1-O1 cluster is > spatially disjoint from the bulk of the suprathreshold data points. Don’t > be mislead by the topographical image, where both the contour lines, and > the 2D placement of the electrodes might reflect a picture that is > different from what you would expect. > > Note, that providing cirange as extra information is not the most > informative. Better would have been to show posclusterslabelmat. This data > field provides for each sample, whether and to which ‘cluster’ it belongs. > > Best wishes, > Jan-Mathijs > > > J.M.Schoffelen, MD PhD > Senior Researcher, VIDI-fellow - PI, language in interaction > Telephone: +31-24-3614793 <+31%2024%20361%204793> > Physical location: room 00.028 > Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands > > > > On 18 Oct 2017, at 23:05, Alea Skwara wrote: > > Hi All, > > I have been conducting a permutation based cluster analysis using the > method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to > examine longitudinal changes in spectral power. Everything runs just fine > and it identifies clusters successfully. > > However, when I examine the results of the analysis, some of the stats > strike me as strange. Electrodes that have the highest F values (I am > comparing across 3 timepoints) do not test as significant, while electrodes > with lower F values do. At first I thought this was because I had a single > "hot" electrode that was not part of a larger cluster, but looking more > closely at the stats, it appears that this is not the case. > > I have tried double-checking my neighbor definition method and electrode > locations to assure that it is not an issue with neighbors being > incorrectly registered. I have also tried adjusting the minimum cluster > size using the minnbchan option to ensure it was not an issue of a cluster > not reaching the minimum extent. Neither of these options seem to have any > impact on the result. > > At this point I'm at a loss. Is there something I'm misunderstanding about > the algorithm? > > Code below and plot below, channel-by-channel stats attached. > > Thanks! > Alea > > One funky thing to note about this code is that I am inputting average > values into the analysis (averaged across time as participants have > slightly differing lengths of data, and within bands as I am running this > analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band > power", not 7 Hz) > > > %% THREE CONDITIONS > %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) > cfg = []; > cfg.channel = {'all'}; > cfg.latency = 'all'; > cfg.frequency = [7 7]; %frequency band indicator > cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing > to average > cfg.method = 'montecarlo'; > cfg.statistic = 'depsamplesFmultivariate'; %three time points > cfg.correctm = 'cluster'; > cfg.clusteralpha = 0.1; %change to .1 because we're computing over > averaged bands and time (losing power) > cfg.clusterstatistic = 'maxsum'; > cfg.minnbchan = 2; > cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise > cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise > cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 > otherwise > cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account > for 500 tests and correction by Bonferroni method. > %specify which electrodes can form clusters > cfg_neighb.method = 'distance'; > cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); > %cfg.neighbours = cfgManish.neighbours; > %cfg = getNeighbours(cfg); % using fixed neighbours > cfg.layout = layout'; > > > % Set up the design matrixfor 3 conditions > subj = 27; %will differ based on condition > > design = zeros(2,3*subj); > for isub = 1:1:subj > design(1,isub) = isub; > design(1,subj+isub) = isub; > design(1,2*subj+isub) = isub; > end > design(2,1:subj) = 1; > design(2,subj+1:2*subj) = 2; > design(2,2*subj+1:3*subj) = 3; > > cfg.design = design; > cfg.uvar = 1; > cfg.ivar = 2; > > > % Run the analysis > [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, > r1t3r_struct); > > > %Plot the results > cfg1 = []; > cfg1.alpha = 0.1; > cfg1.parameter = 'stat'; > %cfg1.zlim = [0 18.1]; > cfg1.layout = layout'; > cfg1.marker = 'o'; > cfg1.subplotsize = [1 1]; > cfg1.saveaspng = 'r1r_betacluster'; > > ft_clusterplot(cfg1, r1r_stat_beta); > > > ***Note non-significant hot spot over left parietal*** > > > > > > > Alea C. Skwara > Graduate Student | Saron Lab > http://mindbrain.ucdavis.edu/labs/Saron > > Center for Mind and Brain | University of California, Davis > 267 Cousteau Place | Davis CA 95616 > Cell: (828) 273-8595 > > > > > > > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > > > > > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > -- Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 -------------- next part -------------- An HTML attachment was scrubbed... URL: From caschera at dis.uniroma1.it Tue Oct 24 13:26:42 2017 From: caschera at dis.uniroma1.it (Stefano Caschera) Date: Tue, 24 Oct 2017 13:26:42 +0200 Subject: [FieldTrip] Problem with OpenMEEG Message-ID: Hi, I am using Matlab 2012b on Windows7 64bit. I have installed OpenMEEG.exe but I had this warning: *Warning! PATH too long installer unable to modify PATH!* So I tried through OpenMEEG.tar.gz, but when I try to follow the tutorial I have this warning: *'om_assemble' is not recognized as an internal or external command, * *operable program or batch file.* I added all OpenMeeg folder with subfolders to the Matlab search path, but If I write *system ('om_assemble') *the answer is the expected one only when I am inside the folder /bin. Any suggestions? Thanks in advance. Thank you, Stefano -- Stefano Caschera, PhD Student *Neuroelectrical Imaging and BCI lab* Fondazione Santa Lucia, IRCCS Via Ardeatina, 306 I-00179, Rome, Italy *Department of Computer, Control, andManagement Engineering "Antonio Ruberti"* Sapienza, University of Rome V. Ariosto, 25 00185, Rome, Italy Tel +39 06 5150 1165 Email: stefano.caschera at uniroma1.it caschera at dis.uniroma1.it s.caschera at hsantalucia.it ___________________________________________________ Mail priva di virus. www.avg.com <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> -------------- next part -------------- An HTML attachment was scrubbed... URL: From caschera at dis.uniroma1.it Tue Oct 24 14:02:33 2017 From: caschera at dis.uniroma1.it (Stefano Caschera) Date: Tue, 24 Oct 2017 14:02:33 +0200 Subject: [FieldTrip] Problem with OpenMEEG Message-ID: Hi, I am using Matlab 2012b on Windows7 64bit. I have installed OpenMEEG.exe but I had this warning: *Warning! PATH too long installer unable to modify PATH!* So I tried through OpenMEEG.tar.gz, but when I try to follow the tutorial I have this warning: *'om_assemble' is not recognized as an internal or external command, * *operable program or batch file.* I added all OpenMeeg folder with subfolders to the Matlab search path, but If I write *system ('om_assemble') *the answer is the expected one only when I am inside the folder /bin. Any suggestions? Thanks in advance. Thank you, Stefano -- Stefano Caschera, PhD Student *Neuroelectrical Imaging and BCI lab* Fondazione Santa Lucia, IRCCS Via Ardeatina, 306 I-00179, Rome, Italy *Department of Computer, Control, andManagement Engineering "Antonio Ruberti"* Sapienza, University of Rome V. Ariosto, 25 00185, Rome, Italy Tel +39 06 5150 1165 Email: stefano.caschera at uniroma1.it caschera at dis.uniroma1.it s.caschera at hsantalucia.it ___________________________________________________ -------------- next part -------------- An HTML attachment was scrubbed... URL: From e.patai at ucl.ac.uk Tue Oct 24 16:57:15 2017 From: e.patai at ucl.ac.uk (Zita Eva Patai) Date: Tue, 24 Oct 2017 15:57:15 +0100 Subject: [FieldTrip] matlab nearest == FT nearest Message-ID: Hi Fters Has anyone else had any issues with using *nearest*? It is now a FT function and a Matlab 2016 one, so it often defaults to the Matlab version, which is unhelpful. If I set the path to the FT version, sometimes Matlab has a spasm. Any input would be helpful, z -- Eva Zita Patai, DPhil Postdoctoral Researcher Institute of Behavioural Neuroscience UCL -------------- next part -------------- An HTML attachment was scrubbed... URL: From Max.Cantor at colorado.edu Thu Oct 26 20:12:45 2017 From: Max.Cantor at colorado.edu (Max Cantor) Date: Thu, 26 Oct 2017 12:12:45 -0600 Subject: [FieldTrip] Comparing/contrasting ft_freqstatistics and eeglab's bootstat Message-ID: Hi, This is not a fieldtrip question per se, but I'm doing something in eeglab and I was wondering if anybody could comment on whether what I'm doing is comparable to fieldtrip's cluster permutation statistic. I'm attempting to create a statistical mask for an event-related spectral perturbation array (specifically a morlet wavelet ersp). The dimensions of the ersp are log-scaled frequency (and where number of cycles increases as frequency increases), samples, and channels. This matrix is the grand average across subjects, and the difference between two conditions. For each subject, each channel, and each condition, the ersps were baselined. In other words, the data are differences in power from baseline and between conditions, in units of decibels. I run the following inputs through bootstat: * [rsignif rbot] = bootstat(permute(g_ersp, [2,1,3]), 'mean(arg1,3);', 'alpha', 0.01, 'dimaccu', 2, 'naccu', 1000);* Where the rsignif output is a freq x 2 array which I use as the statistical mask, g_ersp is the ersp matrix I've been referring to, 'mean(arg1,3)' is the function, alpha is alpha, dimaccu is the dimension to shuffle, and naccu is the number times to reshuffle. This averages across channels (the channels are an ROI so this is what I want), shuffles across samples 1000 times, and tests for significance at alpha = 0.01. It is not testing against a baseline as I understand ft_freqstatistics to do. I use rsignif as an ersp statistical mask, and when I included the baseline vector in bootstat, it failed to mask anything. I think this is because I had baselined the ersp prior to the statistic, so literally any power tested against an empty baseline window was going to be significant. Running it in this way without testing against a baseline, I get "sensible-looking" maskings, but it would be nice to get external confirmation that what I'm doing is methodologically sound, and that I am correctly interpreting my statistic conceptually. I have used ft_freqstatistics in the past and would like to frame this bootstat statistic in a similar manner, which is why I'm asking here. Also, if I am misunderstanding my statistic, advice either on how to properly implement ft_freqstatistics-like cluster permutation statistics in this bootstat function, or alternatively how to convert my ersp matrix in such a way as to be usable with ft_freqstatistics, would be appreciated. Thanks, Max -- Max Cantor Graduate Student Cognitive Neuroscience of Language Lab University of Colorado Boulder -------------- next part -------------- An HTML attachment was scrubbed... URL: From o.jensen at bham.ac.uk Sat Oct 28 09:40:13 2017 From: o.jensen at bham.ac.uk (Ole Jensen) Date: Sat, 28 Oct 2017 08:40:13 +0100 Subject: [FieldTrip] 3 postdoc positions / U of Birmingham Message-ID: <7a17da5d-34c3-ad37-0a78-c67c68bfde5b@bham.ac.uk> Dear all, I have 3 postdoc openings funded by a Wellcome Trust Investigator Award. See here for more information. Applications can be submitted from Nov 1 to 16. ==== Applications are invited for a full-time Postdoctoral Research Fellow to work at the Centre for Human Brain Health, University of Birmingham. The successful candidate will work on a project aimed at studying the role of brain oscillations in relation to perception, attention and memory. This will be done in the group of Prof. Ole Jensen. The methodological focus of the project is MEG recordings and re-analysis of existing animal data. The candidate will be making use of the newly to be installed MEG system at the University of Birmingham and help to develop the laboratory. The candidate will be involved in all stages of designing and conducting experiments, analysing electrophysiological data. The candidate will be expected to prepare the results for high quality academic publications, to present at national and international academic conferences, and to engage in public activities. All research will be conducted in the spirit of open science. Furthermore the candidates will take part in supervision students and contribute to the Centre for Human Brain Health through leadership. === Feel free to write me for details. All the best, Ole -- Prof. Ole Jensen Centre for Human Brain Health, University of Birmingham http://www.neuosc.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From fereshte.ramezani at gmail.com Sun Oct 29 07:40:56 2017 From: fereshte.ramezani at gmail.com (Fereshte) Date: Sun, 29 Oct 2017 10:10:56 +0330 Subject: [FieldTrip] Import your own segmentation for making the head model in FiledTrip Message-ID: Dear Experts, How can one import his own segmentation data ( five labels as ; skull,scalp,GM,WM and CSF) into FiledTrip to make the finite element head model ? Thank you for your attention. Regards, Fereshte -------------- next part -------------- An HTML attachment was scrubbed... URL: From yoland.savriama at helsinki.fi Mon Oct 30 12:47:27 2017 From: yoland.savriama at helsinki.fi (Savriama, Yoland F) Date: Mon, 30 Oct 2017 11:47:27 +0000 Subject: [FieldTrip] How make a heat map on a skull using 3D landmark-based geometric morphometrics? Message-ID: Dear colleagues, I use 3D landmark-based geometric morphometrics for statistical analyses of morphological variation in vertebrate skulls. I would like to know how I could create a heat map of a skull that will show the deviations from the mean shape for a given effect around each landmark. For instance, the patterns of shape changes associated with a given Principal Component that would be mapped on a 3D skull representing the average mean shape. In this case, the visualisations of these deviations will be restricted and associated only with the local landmark coordinates as opposed to a full heat map that would cover the entire skull. In fact, the exact thing I would like to achieve had been done in Figure 4A of this publication: Maga, A. M., Navarro, N., Cunningham, M. L., & Cox, T. C. (2015). Quantitative trait loci affecting the 3D skull shape and size in mouse and prioritization of candidate genes in-silico. Frontiers in physiology, 6. https://doi.org/10.3389/fphys.2015.00092 I would greatly appreciate your feedback on this matter. Best wishes, Yoland SAVRIAMA, PhD Institute of Biotechnology P.O. Box 56 (Viikinkaari 5) FIN-00014 FINLAND -------------- next part -------------- An HTML attachment was scrubbed... URL: From mathilde.bonnefond at inserm.fr Mon Oct 30 17:07:04 2017 From: mathilde.bonnefond at inserm.fr (mathilde.bonnefond at inserm.fr) Date: Mon, 30 Oct 2017 17:07:04 +0100 Subject: [FieldTrip] postdoctoral position on the mechanistic role of brain oscillations In-Reply-To: <230d4c0ddd0a8ac1e161b1dbed847738@inserm.fr> References: <230d4c0ddd0a8ac1e161b1dbed847738@inserm.fr> Message-ID: Dear all, I'm offering a post-doc position on the role of nested oscillations in brain networks communication at the Lyon Neuroscience Research Center (CRNL) starting in February (see attached document). I would be grateful if you could share the attached job offer with interested PhD students and postdocs around you. Thank you very much. Best regards, Mathilde Bonnefond ---- Mathilde Bonnefond, PhD Lyon Neuroscience Research Center (CRNL) Inserm-CNRS-University Lyon 1 Dycog group Bron, France -------------- next part -------------- A non-text attachment was scrubbed... Name: PostDocAdvertisment.pdf Type: application/pdf Size: 232069 bytes Desc: not available URL: From heqing.psy at hotmail.com Tue Oct 31 08:45:06 2017 From: heqing.psy at hotmail.com (He Qing) Date: Tue, 31 Oct 2017 07:45:06 +0000 Subject: [FieldTrip] import Curry dataset into FieldTrip Message-ID: Dear FieldTripist: I am new to FiledTrip, could someone tell me how to read Curry dataset into FieldTrip? When I call the ft_preprocessing or ft_definetrial, it is alway showing errors. ---------------------------------------------- Error using ft_notification (line 340) unsupported header format "curry_dat" Qing He -------------- next part -------------- An HTML attachment was scrubbed... URL: From isac.sehlstedt at psy.gu.se Mon Oct 2 13:03:17 2017 From: isac.sehlstedt at psy.gu.se (Isac Sehlstedt) Date: Mon, 2 Oct 2017 11:03:17 +0000 Subject: [FieldTrip] Follow up question: Computing pca variables (i.e. Latent, and coefficient variables) after ft_componentanalysis (Schoffelen, J.M. (Jan Mathijs)) Message-ID: Dear Jan, I am such a blind hen sometimes. Not the proudest moment of my life. Thank you for helping me find what I was looking for. Very Best, Isac -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 09:33:12 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 07:33:12 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> Message-ID: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 09:39:04 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 07:39:04 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization Message-ID: <6d86574c600e4bc4a114b4d7ea788d18@neuro.gu.se> Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz -------------- next part -------------- An HTML attachment was scrubbed... URL: From sarang at cfin.au.dk Wed Oct 4 09:47:19 2017 From: sarang at cfin.au.dk (Sarang S. Dalal) Date: Wed, 4 Oct 2017 07:47:19 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> Message-ID: <1507103238.2609.8.camel@cfin.au.dk> Hi Bushra, Are you using the common filter approach to use the same spatial filter for your contrasted conditions? http://www.fieldtriptoolbox.org/example/common_filters_in_beamforming (Although that example is focused on DICS and PCC, a similar approach can also be used for LCMV.) This should reduce the likelihood of the 'middle of the head' artifact. If the filters are computed independently, however, then the noise suppression characteristics of each filter can be different, so the contrast would pull out differences in each filter's noise performance. If you are already using this approach, then you can try weight normalization, which is one strategy for removing the noise bias (and usually prevents the 'center of the head' artifact even when no contrast is done). You can run this by adding to your cfg structure: cfg.lcmv.weightnorm = 'nai' - or - cfg.lcmv.weightnorm = 'unitnoisegain'; Each of those methods should give you the same result with a contrast. The NAI is simply scaled against the theoretical noise power (i.e., the noise level = 1), so is often preferred when no contrast is done. Cheers, Sarang On Wed, 2017-10-04 at 07:33 +0000, Bushra Riaz Syeda wrote: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don’t see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 10:19:09 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 08:19:09 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: <1507103238.2609.8.camel@cfin.au.dk> References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> , <1507103238.2609.8.camel@cfin.au.dk> Message-ID: Thank you for you reply I am already using common filter for both conditions and I have also tried weight normalisation with 'unitnoisegain', but it didn't change results. My protocol is bit unusual in a way that electric stimulation to the finger is time locked with 200 ms after the R wave of ECG. As such, the ECG artifact is prominent at -200 ms in my data and I am using ICA to remove it as shown in figure (https://chalmersuniversity.box.com/s/deuzucqovgcfienybk7zisc3r3wjusvu) For lcmv I am using -0.71 to -0.21 sec of prestimulus interval window ( relatively clean baseline without ECG artifact) and 0.01 to 0.51 sec of post stimulus interval window. Could my protocol be the reason of 'center of head artifact', I didnt think it would be the case? Any suggestions to solve this would be highly appreciated. Thank you Kind regards Bushra Riaz ________________________________ From: fieldtrip-bounces at science.ru.nl on behalf of Sarang S. Dalal Sent: Wednesday, October 4, 2017 9:47:19 AM To: fieldtrip at science.ru.nl Subject: Re: [FieldTrip] Artifact in LCMV beamformer source localization Hi Bushra, Are you using the common filter approach to use the same spatial filter for your contrasted conditions? http://www.fieldtriptoolbox.org/example/common_filters_in_beamforming (Although that example is focused on DICS and PCC, a similar approach can also be used for LCMV.) This should reduce the likelihood of the 'middle of the head' artifact. If the filters are computed independently, however, then the noise suppression characteristics of each filter can be different, so the contrast would pull out differences in each filter's noise performance. If you are already using this approach, then you can try weight normalization, which is one strategy for removing the noise bias (and usually prevents the 'center of the head' artifact even when no contrast is done). You can run this by adding to your cfg structure: cfg.lcmv.weightnorm = 'nai' - or - cfg.lcmv.weightnorm = 'unitnoisegain'; Each of those methods should give you the same result with a contrast. The NAI is simply scaled against the theoretical noise power (i.e., the noise level = 1), so is often preferred when no contrast is done. Cheers, Sarang On Wed, 2017-10-04 at 07:33 +0000, Bushra Riaz Syeda wrote: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From na.so.ir at gmail.com Thu Oct 5 12:44:35 2017 From: na.so.ir at gmail.com (Narjes Soltani) Date: Thu, 5 Oct 2017 12:44:35 +0200 Subject: [FieldTrip] Median Filter Message-ID: Dear Fieldtripers, I have a question about the median filter. When I set this filter to 'no' (which is the default value) in ft_preprocessing configuration file and then run the ICA on the output, multiple components are not converged. However if I set the median filter to 'yes', this problem vanishes. Is there any idea why setting the median filter to 'yes' which means keeping the jump artifacts, helps in ICA convergence? By the way, I sent the epoched data to ft_preprocessing which may also be important about how ICA works. Very Best, Narjes -------------- next part -------------- An HTML attachment was scrubbed... URL: From chrichat at hotmail.com Fri Oct 6 15:42:02 2017 From: chrichat at hotmail.com (Christos Chatzichristos) Date: Fri, 6 Oct 2017 13:42:02 +0000 Subject: [FieldTrip] Simulation of sources based on fMRI spatial map. Message-ID: Dear all I am new at the community of field trip (and EEG ) generally. I am a PHD student and I am working on tensor representation for fMRI Blind Source Separation. Currently I am working on a parallel project for fusion of EEG and fMRI (with tensors again). Since there are new methods before going to real data I need some realistic simulated to test my methods on. I went through the tutorials about `Creating a sourcemodel for source-reconstruction of MEG or EEG data' and also the example 'Compute forward simulated data and apply a dipole fit' but still I was wondering if there is a way to create the leadfield matrices for sources simulated for fMRI or even more simply, from single slice sources like the one attached (which has been simulated with SimTB for a spherical brain). I was thinking to perform a simulation similar to this used in [X. Lei et al. 'A parallel framework for simultaneous EEG/fMRI analysis: Methodology and simulation], where they build a concentric three-sphere model around a spherical brain slice similar to the one I attach. Thanks a lot for any possible help. Kind regards Christos Chatzichristos -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Screenshot from 2017-10-06 14-37-38.png Type: image/png Size: 17053 bytes Desc: Screenshot from 2017-10-06 14-37-38.png URL: From aarjona at us.es Fri Oct 6 12:09:48 2017 From: aarjona at us.es (Antonio Arjona Valladares) Date: Fri, 06 Oct 2017 12:09:48 +0200 Subject: [FieldTrip] Fwd: the field cfg.neighbours is required In-Reply-To: References: Message-ID: Hello everyone! This is my query: Im trying to make statistical analyses of time-frequency with two groups of subjects. I have created the file .study with both groups, precompute de channels, and then I would like to apply the 'montecarlo/permutation statistics' with 'cluster correction'. The problem comes when I select all the electrodes and click on 'Plot ERSP(s)'. After a few seconds, the program gives me this answer (image 3). I have been reading tutorials, but nothings appears about this issue. Thank you so much --- Antonio Arjona Valladares Research Technician (BS) Human Psychobiology Lab, Experimental Psychology Department, University of Seville, C/Camilo José Cela s/n, 41018-Sevilla, Spain Tel: +34 954 55 78 00 - 954 55 69 41 -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image 3.png Type: image/png Size: 1067786 bytes Desc: not available URL: From abela.eugenio at gmail.com Fri Oct 6 17:30:50 2017 From: abela.eugenio at gmail.com (Eugenio Abela) Date: Fri, 6 Oct 2017 16:30:50 +0100 Subject: [FieldTrip] Fwd: the field cfg.neighbours is required In-Reply-To: References: Message-ID: <9A6240C6-2D77-4A26-BB76-3DDFD61CD417@gmail.com> Hi Antonio this looks like a question that EEGLAB might also be able to answer (https://sccn.ucsd.edu/wiki/EEGLAB_mailing_lists )? In any case, see here for what cfg.neighbours means /does: http://www.fieldtriptoolbox.org/tutorial/cluster_permutation_timelock http://www.fieldtriptoolbox.org/faq/how_can_i_define_neighbouring_sensors I’m sorry, but I don’t know how you specify the neighbour structure in EEGLAB... Good luck Eugenio On 6 Oct 2017, at 11:09, Antonio Arjona Valladares wrote: Hello everyone! This is my query: Im trying to make statistical analyses of time-frequency with two groups of subjects. I have created the file .study with both groups, precompute de channels, and then I would like to apply the 'montecarlo/permutation statistics' with 'cluster correction'. The problem comes when I select all the electrodes and click on 'Plot ERSP(s)'. After a few seconds, the program gives me this answer (image 3). I have been reading tutorials, but nothings appears about this issue. Thank you so much --- Antonio Arjona Valladares Research Technician (BS) Human Psychobiology Lab, Experimental Psychology Department, University of Seville, C/Camilo José Cela s/n, 41018-Sevilla, Spain Tel: +34 954 55 78 00 - 954 55 69 41 _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image 3.png Type: image/png Size: 231844 bytes Desc: not available URL: From cmhan616 at utexas.edu Tue Oct 10 21:54:57 2017 From: cmhan616 at utexas.edu (Chungmin Han) Date: Tue, 10 Oct 2017 14:54:57 -0500 Subject: [FieldTrip] Question on TFR analysis Message-ID: Dear all, I'm Chungmin, graduate student in UT Austin, and have a question TFR using fieldtrip, my experiment has 2 triggers, first one is Trial onset second trigger is the beginning of task with fixed time when subject trigger to begin task. Thus there's varying time between 1st and 2nd trigger, I need to analyze signal with fixed time window after 2nd trigger but need to take baseline right before 1st trigger because there's movement between 1st and 2nd, is there a way I can set up baseline window in matrix form?? seems TFRanalysis allows single baseline time window settings not for varying time window, I've also looked at redefine_trials, however, in this case, I'll have to put two inputs to calculate TFR, one as baselines only and the other task only. .....if either of the methods is not possible, can I set up single baseline (e.g. resting state data)??? Any advice is appreciated Thank you, Best Chungmin -------------- next part -------------- An HTML attachment was scrubbed... URL: From M.vanEs at donders.ru.nl Wed Oct 11 13:58:14 2017 From: M.vanEs at donders.ru.nl (Es, M.W.J. van (Mats)) Date: Wed, 11 Oct 2017 11:58:14 +0000 Subject: [FieldTrip] Question on TFR analysis (Chungmin Han) Message-ID: <3FC79061C73BEF44A3BEDA5DFC0ADBDF86529C4F@EXPRD99.hosting.ru.nl> Hi Chungmin, You are right, it is not possible to make your contrast in a single step. It is not possible to do a baseline correction in ft_freqbaseline (also holds for ft_singleplotTFR etc.) when the baseline is time locked to a different event than the 'active' condition. In order to get the contrast you should get the TFR for both 'active' and 'baseline' condition, timelocked to the different events (by using ft_redefinetrial). You can then use ft_math to calculate the difference. Good luck, Mats van Es -----Original Message----- From: fieldtrip-bounces at science.ru.nl [mailto:fieldtrip-bounces at science.ru.nl] On Behalf Of fieldtrip-request at science.ru.nl Sent: woensdag 11 oktober 2017 12:00 To: fieldtrip at science.ru.nl Subject: fieldtrip Digest, Vol 83, Issue 6 Send fieldtrip mailing list submissions to fieldtrip at science.ru.nl To subscribe or unsubscribe via the World Wide Web, visit https://mailman.science.ru.nl/mailman/listinfo/fieldtrip or, via email, send a message with subject or body 'help' to fieldtrip-request at science.ru.nl You can reach the person managing the list at fieldtrip-owner at science.ru.nl When replying, please edit your Subject line so it is more specific than "Re: Contents of fieldtrip digest..." Today's Topics: 1. Question on TFR analysis (Chungmin Han) ---------------------------------------------------------------------- Message: 1 Date: Tue, 10 Oct 2017 14:54:57 -0500 From: Chungmin Han To: fieldtrip at science.ru.nl Subject: [FieldTrip] Question on TFR analysis Message-ID: Content-Type: text/plain; charset="utf-8" Dear all, I'm Chungmin, graduate student in UT Austin, and have a question TFR using fieldtrip, my experiment has 2 triggers, first one is Trial onset second trigger is the beginning of task with fixed time when subject trigger to begin task. Thus there's varying time between 1st and 2nd trigger, I need to analyze signal with fixed time window after 2nd trigger but need to take baseline right before 1st trigger because there's movement between 1st and 2nd, is there a way I can set up baseline window in matrix form?? seems TFRanalysis allows single baseline time window settings not for varying time window, I've also looked at redefine_trials, however, in this case, I'll have to put two inputs to calculate TFR, one as baselines only and the other task only. .....if either of the methods is not possible, can I set up single baseline (e.g. resting state data)??? Any advice is appreciated Thank you, Best Chungmin -------------- next part -------------- An HTML attachment was scrubbed... URL: ------------------------------ _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip End of fieldtrip Digest, Vol 83, Issue 6 **************************************** From luca.turella at unitn.it Wed Oct 11 17:06:13 2017 From: luca.turella at unitn.it (Luca Turella) Date: Wed, 11 Oct 2017 17:06:13 +0200 Subject: [FieldTrip] Post-doc in Neural dynamics of Action understanding @ CIMeC University of Trento, Italy Message-ID: A postdoctoral position will be available soon at the Center for Mind/Brain Sciences (CIMeC, http://www.cimec.unitn.it/en) at the University of Trento (Italy). The topic of investigation will cover the neural dynamics underlying action understanding adopting behavioural studies and MEG. The requirements for the candidates are the following. Candidates must have a Ph.D. degree in a field related to Cognitive Neuroscience or related areas. Previous experience in behavioural/kinematic studies and/or EEG/MEG data analysis is required. The ideal candidate should have good programming skills in Matlab, preferably related to EEG/MEG data analysis (Fieldtrip). Knowledge of kinematic data analysis is also a plus. Knowledge of Italian language is not required. The salary will be proportional to the level of experience and the starting date of the appointment is negotiable, but within the next 6 months. Applications will be considered until the position is filled. The contract will have a duration of 2 years. Applications should be sent to luca.turella at unitn.it, including complete CV, statement of research interests, and contact details of two referees. Potential candidates are also encouraged to send me directly informal inquires *luca.turella at unitn.it *. CIMeC offers an international and vibrant research setting with access to state-of-the-art neuroimaging methodologies, including a research-only MR scanner, MEG, EEG and TMS, as well as behavioural, eye tracking and motion tracking laboratories. English is the official language of the CIMeC, where a large proportion of the faculty, post-docs and students come from a wide range of countries outside of Italy. The University of Trento consistently ranks as a top Italian university in both national Research Assessment Evaluations (RAE) and University Surveys. In the latest RAE, the University of Trento as a whole ranks 2nd among medium-sized universities. -- Luca Turella, PhD Assistant Professor CIMeC - Center for Mind/Brain Sciences University of Trento Mattarello (TN), Via Delle Regole 101 Tel.+39 0461-28 3098 http://www.unitn.it/cimec Legal Disclaimer This electronic message contains information that is confidential. The information is intended for the use of the addressee only. If you are not the addressee we would appreciate your notification in this respect. Please note that any disclosure, copy, distribution or use of the contents of this message is prohibited and may be unlawful. Avvertenza legale Questo messaggio Email contiene informazioni confidenziali riservate ai soli destinatari. Qualora veniate in possesso di tali informazioni senza essere definito come destinatario vi reghiamo di leggere le seguenti note. Ogni apertura, copia, distribuzione del contenuto del messaggio e dei suoi allegati è proibito e potrebbe violare le presenti leggi. -------------- next part -------------- An HTML attachment was scrubbed... URL: From m.stoica at uke.de Wed Oct 11 21:54:41 2017 From: m.stoica at uke.de (Mircea Stoica) Date: Wed, 11 Oct 2017 21:54:41 +0200 Subject: [FieldTrip] Sharp transition in multitaper coherence when moving from one taper to two Message-ID: Dear fieldtrippers, I'm interested in coherence analysis with mtmconvol and multitapers. My frequencies of interest are spaced logarithmically (steps of 1/16 octaves) and I adjust the time window and number of tapers to yield a constant frequency smoothing of 0.75 octaves (half that for cfg.tapsmofrq of course). This results in an ugly step-like transition of coherence when moving from one taper to two, as you can see in the following image. https://photos.app.goo.gl/5uWy05RzZxaA9gci1 This sharp transition is not visible in the power output. I calculated the actual frequency smoothing as fw = (K + 1) ./ cfg.t_ftimwin / 2 and in a semilogy plot it gives this: https://photos.app.goo.gl/xQsrY1jr8L0qLLDm2 And the number of tapers for each frequency: https://photos.app.goo.gl/LqaeUVW6ymBWEKEx2 There are no discontinuities in the frequency smoothing. Any ideas what could be the cause? Or is there perhaps nothing to be done about it? Best, Mircea -- _____________________________________________________________________ Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg | www.uke.de Vorstandsmitglieder: Prof. Dr. Burkhard Göke (Vorsitzender), Prof. Dr. Dr. Uwe Koch-Gromus, Joachim Prölß, Martina Saurin (komm.) _____________________________________________________________________ SAVE PAPER - THINK BEFORE PRINTING -------------- next part -------------- An HTML attachment was scrubbed... URL: From stephen.whitmarsh at gmail.com Fri Oct 13 17:29:35 2017 From: stephen.whitmarsh at gmail.com (Stephen Whitmarsh) Date: Fri, 13 Oct 2017 17:29:35 +0200 Subject: [FieldTrip] NAI on trial-by-trail power estimates Message-ID: Dear all, I am trying to optimize my beamformer power estimates, which are not bad, but I want to see if I can improve them. I do not seem to have a depth-bias, and am testing within-subject between-condition contrasts, for which one does not need to correct for noise differences. However, my analysis does involve trial-by-trial analysis, and might be susceptible to noise differences over time. I was wondering whether it would make sense to have trial-by-trial corrections using the Neural Activation Index (NAI), as described in the beamformer tutorial, i.e. by dividing my single-trial power estimates by single-trial noise estimates. Has anyone tried using either the NAI on a trial-by-trial basis? Secondly, does this even make sense to you? Cheers, Stephen -------------- next part -------------- An HTML attachment was scrubbed... URL: From yuxxx955 at umn.edu Mon Oct 16 23:26:30 2017 From: yuxxx955 at umn.edu (Kai Yu) Date: Mon, 16 Oct 2017 16:26:30 -0500 Subject: [FieldTrip] How to understand the cross-correlogram and its feature comparison against the one obtained by shuffling the trials Message-ID: Hi, I have been using your FieldTrip Intracranial Spike Analysis package for a while. However, I am still confused by the statement listed on your website. "Cross-correlations between neurons can either arise because of common, time-locked fluctuations in the firing rate (Brody et al., 1999). These correlations are invariant to a change in the order of trials. The shuffling of trials in ft_spike_xcorr always pertains to two subsequent trials, in order to avoid an influence of slow changes in the firing rate across trials. We refer to this cross-correlogram that is obtained under a permutation of subsequent trials as the 'shift-predictor' cross-correlogram. If the observed features of the cross-correlogram that are not present in the shift-predictor cross-correlogram, then this indicates that they arise because of induced synchronous activity." For the last sentence, would you please explained more about the "induced synchronous activity"? In my own understanding, this is the activity after the task or external stimulation. But if I want to know the synchronous or de-synchronous activity during the stimulation, how can I get the evidence from these cross-correlations and also the jpsth figures? Thank you so much! Best, -Kai -- Kai Yu Biomedical Functional Imaging and Neuroengineering Laboratory Department of Biomedical Engineering University of Minnesota - Twin Cities -------------- next part -------------- An HTML attachment was scrubbed... URL: From acskwara at ucdavis.edu Wed Oct 18 23:05:45 2017 From: acskwara at ucdavis.edu (Alea Skwara) Date: Wed, 18 Oct 2017 14:05:45 -0700 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) Message-ID: Hi All, I have been conducting a permutation based cluster analysis using the method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to examine longitudinal changes in spectral power. Everything runs just fine and it identifies clusters successfully. However, when I examine the results of the analysis, some of the stats strike me as strange. Electrodes that have the highest F values (I am comparing across 3 timepoints) do not test as significant, while electrodes with lower F values do. At first I thought this was because I had a single "hot" electrode that was not part of a larger cluster, but looking more closely at the stats, it appears that this is not the case. I have tried double-checking my neighbor definition method and electrode locations to assure that it is not an issue with neighbors being incorrectly registered. I have also tried adjusting the minimum cluster size using the minnbchan option to ensure it was not an issue of a cluster not reaching the minimum extent. Neither of these options seem to have any impact on the result. At this point I'm at a loss. Is there something I'm misunderstanding about the algorithm? Code below and plot below, channel-by-channel stats attached. Thanks! Alea One funky thing to note about this code is that I am inputting average values into the analysis (averaged across time as participants have slightly differing lengths of data, and within bands as I am running this analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band power", not 7 Hz) %% THREE CONDITIONS %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) cfg = []; cfg.channel = {'all'}; cfg.latency = 'all'; cfg.frequency = [7 7]; %frequency band indicator cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing to average cfg.method = 'montecarlo'; cfg.statistic = 'depsamplesFmultivariate'; %three time points cfg.correctm = 'cluster'; cfg.clusteralpha = 0.1; %change to .1 because we're computing over averaged bands and time (losing power) cfg.clusterstatistic = 'maxsum'; cfg.minnbchan = 2; cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 otherwise cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account for 500 tests and correction by Bonferroni method. %specify which electrodes can form clusters cfg_neighb.method = 'distance'; cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); %cfg.neighbours = cfgManish.neighbours; %cfg = getNeighbours(cfg); % using fixed neighbours cfg.layout = layout'; % Set up the design matrixfor 3 conditions subj = 27; %will differ based on condition design = zeros(2,3*subj); for isub = 1:1:subj design(1,isub) = isub; design(1,subj+isub) = isub; design(1,2*subj+isub) = isub; end design(2,1:subj) = 1; design(2,subj+1:2*subj) = 2; design(2,2*subj+1:3*subj) = 3; cfg.design = design; cfg.uvar = 1; cfg.ivar = 2; % Run the analysis [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, r1t3r_struct); %Plot the results cfg1 = []; cfg1.alpha = 0.1; cfg1.parameter = 'stat'; %cfg1.zlim = [0 18.1]; cfg1.layout = layout'; cfg1.marker = 'o'; cfg1.subplotsize = [1 1]; cfg1.saveaspng = 'r1r_betacluster'; ft_clusterplot(cfg1, r1r_stat_beta); ***Note non-significant hot spot over left parietal*** [image: Inline image 2] Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: r1r_betacluster_fig1.png Type: image/png Size: 49573 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: clusterstats_beta.xlsx Type: application/vnd.openxmlformats-officedocument.spreadsheetml.sheet Size: 11196 bytes Desc: not available URL: From bioeng.yoosofzadeh at gmail.com Thu Oct 19 05:34:09 2017 From: bioeng.yoosofzadeh at gmail.com (Vahab Yousofzadeh) Date: Wed, 18 Oct 2017 23:34:09 -0400 Subject: [FieldTrip] Postdoctoral position at Ulster on Dementia and neuroimaging Message-ID: Hi everyone, There is a postdoc position at Ulster Univerity with Dr. KongFatt Wong-Lin and Prof. Tony Bjourson on dementia and neuroimaging. Check it out: https://goo.gl/4YexD5 Cheers, Vahab From jan.schoffelen at donders.ru.nl Thu Oct 19 08:19:35 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Thu, 19 Oct 2017 06:19:35 +0000 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) In-Reply-To: References: Message-ID: Hi Alea, I think that it all looks OK. I guess that the P5-PO1-O1 cluster is spatially disjoint from the bulk of the suprathreshold data points. Don’t be mislead by the topographical image, where both the contour lines, and the 2D placement of the electrodes might reflect a picture that is different from what you would expect. Note, that providing cirange as extra information is not the most informative. Better would have been to show posclusterslabelmat. This data field provides for each sample, whether and to which ‘cluster’ it belongs. Best wishes, Jan-Mathijs J.M.Schoffelen, MD PhD Senior Researcher, VIDI-fellow - PI, language in interaction Telephone: +31-24-3614793 Physical location: room 00.028 Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands On 18 Oct 2017, at 23:05, Alea Skwara > wrote: Hi All, I have been conducting a permutation based cluster analysis using the method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to examine longitudinal changes in spectral power. Everything runs just fine and it identifies clusters successfully. However, when I examine the results of the analysis, some of the stats strike me as strange. Electrodes that have the highest F values (I am comparing across 3 timepoints) do not test as significant, while electrodes with lower F values do. At first I thought this was because I had a single "hot" electrode that was not part of a larger cluster, but looking more closely at the stats, it appears that this is not the case. I have tried double-checking my neighbor definition method and electrode locations to assure that it is not an issue with neighbors being incorrectly registered. I have also tried adjusting the minimum cluster size using the minnbchan option to ensure it was not an issue of a cluster not reaching the minimum extent. Neither of these options seem to have any impact on the result. At this point I'm at a loss. Is there something I'm misunderstanding about the algorithm? Code below and plot below, channel-by-channel stats attached. Thanks! Alea One funky thing to note about this code is that I am inputting average values into the analysis (averaged across time as participants have slightly differing lengths of data, and within bands as I am running this analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band power", not 7 Hz) %% THREE CONDITIONS %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) cfg = []; cfg.channel = {'all'}; cfg.latency = 'all'; cfg.frequency = [7 7]; %frequency band indicator cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing to average cfg.method = 'montecarlo'; cfg.statistic = 'depsamplesFmultivariate'; %three time points cfg.correctm = 'cluster'; cfg.clusteralpha = 0.1; %change to .1 because we're computing over averaged bands and time (losing power) cfg.clusterstatistic = 'maxsum'; cfg.minnbchan = 2; cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 otherwise cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account for 500 tests and correction by Bonferroni method. %specify which electrodes can form clusters cfg_neighb.method = 'distance'; cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); %cfg.neighbours = cfgManish.neighbours; %cfg = getNeighbours(cfg); % using fixed neighbours cfg.layout = layout'; % Set up the design matrixfor 3 conditions subj = 27; %will differ based on condition design = zeros(2,3*subj); for isub = 1:1:subj design(1,isub) = isub; design(1,subj+isub) = isub; design(1,2*subj+isub) = isub; end design(2,1:subj) = 1; design(2,subj+1:2*subj) = 2; design(2,2*subj+1:3*subj) = 3; cfg.design = design; cfg.uvar = 1; cfg.ivar = 2; % Run the analysis [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, r1t3r_struct); %Plot the results cfg1 = []; cfg1.alpha = 0.1; cfg1.parameter = 'stat'; %cfg1.zlim = [0 18.1]; cfg1.layout = layout'; cfg1.marker = 'o'; cfg1.subplotsize = [1 1]; cfg1.saveaspng = 'r1r_betacluster'; ft_clusterplot(cfg1, r1r_stat_beta); ***Note non-significant hot spot over left parietal*** Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From maximilien.chaumon at gmail.com Thu Oct 19 11:18:51 2017 From: maximilien.chaumon at gmail.com (Maximilien Chaumon) Date: Thu, 19 Oct 2017 09:18:51 +0000 Subject: [FieldTrip] memory leak in ft_selectdata Message-ID: Dear all, I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion cfg = []; cfg.parameter = 'logmpowspctrm'; cfg.comment = 'no'; cfg.xlim = [20 21]; cfg.zlim = 'maxmin'; cfg.layout = 'neuromag306all.lay'; cfg.colorbar = 'no'; big_spec_grad.dimord = 'chan_freq'; ft_topoplotER(cfg,big_spec_grad); This used to work fine until a last update this week from a version I suspect dated back to 03/17. Now today, the ft_topoplotER line eats up a lot of memory the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. x = x(:,selindx,:,:,:,:); The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... Any help is much appreciated. Thanks! Max -------------- next part -------------- An HTML attachment was scrubbed... URL: From adonay.s.nunes at gmail.com Thu Oct 19 23:41:53 2017 From: adonay.s.nunes at gmail.com (A Nunes) Date: Thu, 19 Oct 2017 14:41:53 -0700 Subject: [FieldTrip] SAM gareth method without depth biased weights Message-ID: Hi community, I have an issue with Fieltrip SAM beamformer 'gareth' method. As a sanity check I look that the BF weight distributions are biased towards the center. I check this distribution with the leadfields and with the BF weights from LCMV and SAMs methods. Gareth method does not have this distribution, the weight sizes are scattered across the brain and the norm of the weights can have up to 11 orders of magnitude of difference. First I make sure that the 1/Leadfield are also biased towards the center: - for every LF I take the norm of every column and 1/mean of the norms is the value. - the code is: LF_norm = []; for s= 1:size(LF, 1) n1 = norm(squeeze(LF(s, 1, :))); n2 = norm(squeeze(LF(s, 2, :))); n3 = norm(squeeze(LF(s, 3, :))); LF_norm(s) =1/mean([n1, n2, n3]); end When interpolated to the mri and plotted looks fine. Then I run SAM robert method: cfg = []; cfg.method = 'sam'; cfg.grid = lf; cfg.vol = headmodel; cfg.grad = data.grad; cfg.fixedori = 'robert'; cfg.keepfilter = 'yes'; cfg.keepori = 'yes'; cfg.projectnoise = 'yes'; cfg.senstype = 'MEG'; SAM1_filter = ft_sourceanalysis(cfg, timelock); and take the norm for every BF weight: SAM1_filter_norm= cellfun(@(X) norm(X),SAM1_filter.avg.filter(SAM1_filter.inside)); and then I interpolate it to the mri for visualization. The results look fine, weights biased towards the center and the min and max norm weight are two orders of magnitude difference. The mean of the absolute of the weights norm is in the order of 10e9 and the smallest and highest weight norm are to the order of e8 and e10 If I do the same as the last step but changing to: cfg.fixedori = 'gareth'; Then the weights are not biased towards the center. The mean of the absolute of the weights norm is in the order of 10e16 and the smallest and highest weight norm are to the order e8 and e19, respectively. Thus, gareth method does not have a center biased weight distribution and the max and miminum norm weights have 11 orders of magnitude of difference, meaning that these orders are reflected in the amplitude of the reconstructed time series. To make it handy the difference between gareth and robert methods are: L is the leadfield for a given source. 'Gareth' Y1 = L' * inv_cov * L; Y2 = L' * (inv_cov * inv_cov) * L; [U,S] = eig(Y2,Y1); 'Robert' [U,S] = svd(real(pinv(L' * inv_cov * L))); Then both methods compute 2 of the orientations and choose the salient one: ori1 = U(:,1); ori1 = ori1/norm(ori1); ori2 = U(:,2); ori2 = ori2/norm(ori2); L1 = L * ori1; L2 = L * ori2; if (norm(L1)/norm(L2)) < 1e-6 opt_vox_or = ori2; else opt_vox_or = ori1; end The difference comes from the way orientations are computed, after the weights calculation is the same for both. I tried with different units, i.e. meters, cm, mm but this abnormalities is still present. Does anybody know what can be the cause? Thanks Adonay -------------- next part -------------- An HTML attachment was scrubbed... URL: From jan.schoffelen at donders.ru.nl Fri Oct 20 09:07:10 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Fri, 20 Oct 2017 07:07:10 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: Hi Max, >From your (otherwise clear) description, I am not sure whether your issue is due to the actual data selection itself. Rather, my first thought would be that something goes amiss with recursive piling up of large amounts of irrelevant (for the plotting at least) metadata, that gets assembled while iterating through subsequent interactive plotting steps. It’s kind of a long story, which I am not going to type down here, but to check whether I am thinking into the right direction, could you see whether your problem persists if you call ft_topoplotER in the following way? ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); Thanks, Jan-Mathijs > On 19 Oct 2017, at 11:18, Maximilien Chaumon wrote: > > Dear all, > > I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion > cfg = []; > cfg.parameter = 'logmpowspctrm'; > cfg.comment = 'no'; > cfg.xlim = [20 21]; > cfg.zlim = 'maxmin'; > cfg.layout = 'neuromag306all.lay'; > cfg.colorbar = 'no'; > big_spec_grad.dimord = 'chan_freq'; > ft_topoplotER(cfg,big_spec_grad); > > This used to work fine until a last update this week from a version I suspect dated back to 03/17. > Now today, the ft_topoplotER line eats up a lot of memory > the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB > the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB > the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB > > This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. > > I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. > x = x(:,selindx,:,:,:,:); > The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). > Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... > > Any help is much appreciated. > Thanks! > Max > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip From mibanks at wisc.edu Fri Oct 20 16:16:54 2017 From: mibanks at wisc.edu (MATTHEW I BANKS) Date: Fri, 20 Oct 2017 14:16:54 +0000 Subject: [FieldTrip] Variance estimates for connectivity measures Message-ID: Greetings. Is it possible to use jackknife to get variance estimates for Granger, partial directed coherence and directed transfer function? I use it for wPLI, but cannot figure out how to do it for these other measures. -Matt Banks ____________________________ Matthew I. Banks, Ph.D. Associate Professor Department of Anesthesiology University of Wisconsin 1300 University Avenue, Room 4605 Madison, WI 53706 office tel. (608)261-1143 lab tel. (608)263-6662 fax (608)263-2592 http://anesthesia.wisc.edu/index.php/Banks_Laboratory http://ntp.neuroscience.wisc.edu/banks.htm -------------- next part -------------- An HTML attachment was scrubbed... URL: From maximilien.chaumon at gmail.com Fri Oct 20 16:30:19 2017 From: maximilien.chaumon at gmail.com (Maximilien Chaumon) Date: Fri, 20 Oct 2017 14:30:19 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: Perfect! that fixed it, many thanks! Max Le ven. 20 oct. 2017 à 09:35, Schoffelen, J.M. (Jan Mathijs) < jan.schoffelen at donders.ru.nl> a écrit : > Hi Max, > > From your (otherwise clear) description, I am not sure whether your issue > is due to the actual data selection itself. Rather, my first thought would > be that something goes amiss with recursive piling up of large amounts of > irrelevant (for the plotting at least) metadata, that gets assembled while > iterating through subsequent interactive plotting steps. > It’s kind of a long story, which I am not going to type down here, but to > check whether I am thinking into the right direction, could you see whether > your problem persists if you call ft_topoplotER in the following way? > > > ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); > > Thanks, > > Jan-Mathijs > > > > > On 19 Oct 2017, at 11:18, Maximilien Chaumon < > maximilien.chaumon at gmail.com> wrote: > > > > Dear all, > > > > I'm doing topoplots of rather large spectrum data focused on a tiny > frequency portion > > cfg = []; > > cfg.parameter = 'logmpowspctrm'; > > cfg.comment = 'no'; > > cfg.xlim = [20 21]; > > cfg.zlim = 'maxmin'; > > cfg.layout = 'neuromag306all.lay'; > > cfg.colorbar = 'no'; > > big_spec_grad.dimord = 'chan_freq'; > > ft_topoplotER(cfg,big_spec_grad); > > > > This used to work fine until a last update this week from a version I > suspect dated back to 03/17. > > Now today, the ft_topoplotER line eats up a lot of memory > > the call to "ft_prepare_layout" took 0 seconds and required the > additional allocation of an estimated 0 MB > > the call to "ft_selectdata" took 0 seconds and required the additional > allocation of an estimated 1283 MB > > the call to "ft_topoplotER" took 1 seconds and required the additional > allocation of an estimated 1283 MB > > > > This gets more dramatic the second time this part of the script runs > (with 2500MB added for reach topoplot) and then crashes on the next > iteration with all of 30GB RAM eaten up by this. > > > > I traced down the memory leak to ft_selectdata line 1281 in a > subfunction called cellmatselect. > > x = x(:,selindx,:,:,:,:); > > The memory usage jumps every time this dimension gets selected. (I admit > I do have a lot of frequencies, over 13000, but I need them for my > application). > > Memory comes back to a normal level when the figure is closed, but as I > said, execution with another dataset eats up double the amount of memory > and I don't know how to get out of this... > > > > Any help is much appreciated. > > Thanks! > > Max > > _______________________________________________ > > fieldtrip mailing list > > fieldtrip at donders.ru.nl > > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From jan.schoffelen at donders.ru.nl Fri Oct 20 16:40:34 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Fri, 20 Oct 2017 14:40:34 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: <68BED22E-0754-4991-B4D5-4F9EAE8536CE@donders.ru.nl> You’re welcome. Yet, I wouldn’t say it’s fixed, but at least you have found a way around your problem. Best wishes, Jan-Mathijs On 20 Oct 2017, at 16:30, Maximilien Chaumon > wrote: Perfect! that fixed it, many thanks! Max Le ven. 20 oct. 2017 à 09:35, Schoffelen, J.M. (Jan Mathijs) > a écrit : Hi Max, From your (otherwise clear) description, I am not sure whether your issue is due to the actual data selection itself. Rather, my first thought would be that something goes amiss with recursive piling up of large amounts of irrelevant (for the plotting at least) metadata, that gets assembled while iterating through subsequent interactive plotting steps. It’s kind of a long story, which I am not going to type down here, but to check whether I am thinking into the right direction, could you see whether your problem persists if you call ft_topoplotER in the following way? ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); Thanks, Jan-Mathijs > On 19 Oct 2017, at 11:18, Maximilien Chaumon > wrote: > > Dear all, > > I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion > cfg = []; > cfg.parameter = 'logmpowspctrm'; > cfg.comment = 'no'; > cfg.xlim = [20 21]; > cfg.zlim = 'maxmin'; > cfg.layout = 'neuromag306all.lay'; > cfg.colorbar = 'no'; > big_spec_grad.dimord = 'chan_freq'; > ft_topoplotER(cfg,big_spec_grad); > > This used to work fine until a last update this week from a version I suspect dated back to 03/17. > Now today, the ft_topoplotER line eats up a lot of memory > the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB > the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB > the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB > > This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. > > I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. > x = x(:,selindx,:,:,:,:); > The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). > Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... > > Any help is much appreciated. > Thanks! > Max > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From acskwara at ucdavis.edu Fri Oct 20 20:59:49 2017 From: acskwara at ucdavis.edu (Alea Skwara) Date: Fri, 20 Oct 2017 11:59:49 -0700 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) In-Reply-To: References: Message-ID: Hi J.M., Thanks so much for your reply. The tip to not trust the topographic maps too much helped in my sleuthing. I realized that adjusting the minnbchan parameter actually *does* result in the topographic changes I thought I should see if I was understanding the algorithm correctly (and that I just wasn't seeing these changes in the map that I was using as a quick-check, but that they were present in the stats themselves). Problem solved! Thank you! Alea On Wed, Oct 18, 2017 at 11:19 PM, Schoffelen, J.M. (Jan Mathijs) < jan.schoffelen at donders.ru.nl> wrote: > Hi Alea, > > I think that it all looks OK. I guess that the P5-PO1-O1 cluster is > spatially disjoint from the bulk of the suprathreshold data points. Don’t > be mislead by the topographical image, where both the contour lines, and > the 2D placement of the electrodes might reflect a picture that is > different from what you would expect. > > Note, that providing cirange as extra information is not the most > informative. Better would have been to show posclusterslabelmat. This data > field provides for each sample, whether and to which ‘cluster’ it belongs. > > Best wishes, > Jan-Mathijs > > > J.M.Schoffelen, MD PhD > Senior Researcher, VIDI-fellow - PI, language in interaction > Telephone: +31-24-3614793 <+31%2024%20361%204793> > Physical location: room 00.028 > Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands > > > > On 18 Oct 2017, at 23:05, Alea Skwara wrote: > > Hi All, > > I have been conducting a permutation based cluster analysis using the > method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to > examine longitudinal changes in spectral power. Everything runs just fine > and it identifies clusters successfully. > > However, when I examine the results of the analysis, some of the stats > strike me as strange. Electrodes that have the highest F values (I am > comparing across 3 timepoints) do not test as significant, while electrodes > with lower F values do. At first I thought this was because I had a single > "hot" electrode that was not part of a larger cluster, but looking more > closely at the stats, it appears that this is not the case. > > I have tried double-checking my neighbor definition method and electrode > locations to assure that it is not an issue with neighbors being > incorrectly registered. I have also tried adjusting the minimum cluster > size using the minnbchan option to ensure it was not an issue of a cluster > not reaching the minimum extent. Neither of these options seem to have any > impact on the result. > > At this point I'm at a loss. Is there something I'm misunderstanding about > the algorithm? > > Code below and plot below, channel-by-channel stats attached. > > Thanks! > Alea > > One funky thing to note about this code is that I am inputting average > values into the analysis (averaged across time as participants have > slightly differing lengths of data, and within bands as I am running this > analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band > power", not 7 Hz) > > > %% THREE CONDITIONS > %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) > cfg = []; > cfg.channel = {'all'}; > cfg.latency = 'all'; > cfg.frequency = [7 7]; %frequency band indicator > cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing > to average > cfg.method = 'montecarlo'; > cfg.statistic = 'depsamplesFmultivariate'; %three time points > cfg.correctm = 'cluster'; > cfg.clusteralpha = 0.1; %change to .1 because we're computing over > averaged bands and time (losing power) > cfg.clusterstatistic = 'maxsum'; > cfg.minnbchan = 2; > cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise > cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise > cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 > otherwise > cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account > for 500 tests and correction by Bonferroni method. > %specify which electrodes can form clusters > cfg_neighb.method = 'distance'; > cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); > %cfg.neighbours = cfgManish.neighbours; > %cfg = getNeighbours(cfg); % using fixed neighbours > cfg.layout = layout'; > > > % Set up the design matrixfor 3 conditions > subj = 27; %will differ based on condition > > design = zeros(2,3*subj); > for isub = 1:1:subj > design(1,isub) = isub; > design(1,subj+isub) = isub; > design(1,2*subj+isub) = isub; > end > design(2,1:subj) = 1; > design(2,subj+1:2*subj) = 2; > design(2,2*subj+1:3*subj) = 3; > > cfg.design = design; > cfg.uvar = 1; > cfg.ivar = 2; > > > % Run the analysis > [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, > r1t3r_struct); > > > %Plot the results > cfg1 = []; > cfg1.alpha = 0.1; > cfg1.parameter = 'stat'; > %cfg1.zlim = [0 18.1]; > cfg1.layout = layout'; > cfg1.marker = 'o'; > cfg1.subplotsize = [1 1]; > cfg1.saveaspng = 'r1r_betacluster'; > > ft_clusterplot(cfg1, r1r_stat_beta); > > > ***Note non-significant hot spot over left parietal*** > > > > > > > Alea C. Skwara > Graduate Student | Saron Lab > http://mindbrain.ucdavis.edu/labs/Saron > > Center for Mind and Brain | University of California, Davis > 267 Cousteau Place | Davis CA 95616 > Cell: (828) 273-8595 > > > > > > > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > > > > > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > -- Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 -------------- next part -------------- An HTML attachment was scrubbed... URL: From caschera at dis.uniroma1.it Tue Oct 24 13:26:42 2017 From: caschera at dis.uniroma1.it (Stefano Caschera) Date: Tue, 24 Oct 2017 13:26:42 +0200 Subject: [FieldTrip] Problem with OpenMEEG Message-ID: Hi, I am using Matlab 2012b on Windows7 64bit. I have installed OpenMEEG.exe but I had this warning: *Warning! PATH too long installer unable to modify PATH!* So I tried through OpenMEEG.tar.gz, but when I try to follow the tutorial I have this warning: *'om_assemble' is not recognized as an internal or external command, * *operable program or batch file.* I added all OpenMeeg folder with subfolders to the Matlab search path, but If I write *system ('om_assemble') *the answer is the expected one only when I am inside the folder /bin. Any suggestions? Thanks in advance. Thank you, Stefano -- Stefano Caschera, PhD Student *Neuroelectrical Imaging and BCI lab* Fondazione Santa Lucia, IRCCS Via Ardeatina, 306 I-00179, Rome, Italy *Department of Computer, Control, andManagement Engineering "Antonio Ruberti"* Sapienza, University of Rome V. Ariosto, 25 00185, Rome, Italy Tel +39 06 5150 1165 Email: stefano.caschera at uniroma1.it caschera at dis.uniroma1.it s.caschera at hsantalucia.it ___________________________________________________ Mail priva di virus. www.avg.com <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> -------------- next part -------------- An HTML attachment was scrubbed... URL: From caschera at dis.uniroma1.it Tue Oct 24 14:02:33 2017 From: caschera at dis.uniroma1.it (Stefano Caschera) Date: Tue, 24 Oct 2017 14:02:33 +0200 Subject: [FieldTrip] Problem with OpenMEEG Message-ID: Hi, I am using Matlab 2012b on Windows7 64bit. I have installed OpenMEEG.exe but I had this warning: *Warning! PATH too long installer unable to modify PATH!* So I tried through OpenMEEG.tar.gz, but when I try to follow the tutorial I have this warning: *'om_assemble' is not recognized as an internal or external command, * *operable program or batch file.* I added all OpenMeeg folder with subfolders to the Matlab search path, but If I write *system ('om_assemble') *the answer is the expected one only when I am inside the folder /bin. Any suggestions? Thanks in advance. Thank you, Stefano -- Stefano Caschera, PhD Student *Neuroelectrical Imaging and BCI lab* Fondazione Santa Lucia, IRCCS Via Ardeatina, 306 I-00179, Rome, Italy *Department of Computer, Control, andManagement Engineering "Antonio Ruberti"* Sapienza, University of Rome V. Ariosto, 25 00185, Rome, Italy Tel +39 06 5150 1165 Email: stefano.caschera at uniroma1.it caschera at dis.uniroma1.it s.caschera at hsantalucia.it ___________________________________________________ -------------- next part -------------- An HTML attachment was scrubbed... URL: From e.patai at ucl.ac.uk Tue Oct 24 16:57:15 2017 From: e.patai at ucl.ac.uk (Zita Eva Patai) Date: Tue, 24 Oct 2017 15:57:15 +0100 Subject: [FieldTrip] matlab nearest == FT nearest Message-ID: Hi Fters Has anyone else had any issues with using *nearest*? It is now a FT function and a Matlab 2016 one, so it often defaults to the Matlab version, which is unhelpful. If I set the path to the FT version, sometimes Matlab has a spasm. Any input would be helpful, z -- Eva Zita Patai, DPhil Postdoctoral Researcher Institute of Behavioural Neuroscience UCL -------------- next part -------------- An HTML attachment was scrubbed... URL: From Max.Cantor at colorado.edu Thu Oct 26 20:12:45 2017 From: Max.Cantor at colorado.edu (Max Cantor) Date: Thu, 26 Oct 2017 12:12:45 -0600 Subject: [FieldTrip] Comparing/contrasting ft_freqstatistics and eeglab's bootstat Message-ID: Hi, This is not a fieldtrip question per se, but I'm doing something in eeglab and I was wondering if anybody could comment on whether what I'm doing is comparable to fieldtrip's cluster permutation statistic. I'm attempting to create a statistical mask for an event-related spectral perturbation array (specifically a morlet wavelet ersp). The dimensions of the ersp are log-scaled frequency (and where number of cycles increases as frequency increases), samples, and channels. This matrix is the grand average across subjects, and the difference between two conditions. For each subject, each channel, and each condition, the ersps were baselined. In other words, the data are differences in power from baseline and between conditions, in units of decibels. I run the following inputs through bootstat: * [rsignif rbot] = bootstat(permute(g_ersp, [2,1,3]), 'mean(arg1,3);', 'alpha', 0.01, 'dimaccu', 2, 'naccu', 1000);* Where the rsignif output is a freq x 2 array which I use as the statistical mask, g_ersp is the ersp matrix I've been referring to, 'mean(arg1,3)' is the function, alpha is alpha, dimaccu is the dimension to shuffle, and naccu is the number times to reshuffle. This averages across channels (the channels are an ROI so this is what I want), shuffles across samples 1000 times, and tests for significance at alpha = 0.01. It is not testing against a baseline as I understand ft_freqstatistics to do. I use rsignif as an ersp statistical mask, and when I included the baseline vector in bootstat, it failed to mask anything. I think this is because I had baselined the ersp prior to the statistic, so literally any power tested against an empty baseline window was going to be significant. Running it in this way without testing against a baseline, I get "sensible-looking" maskings, but it would be nice to get external confirmation that what I'm doing is methodologically sound, and that I am correctly interpreting my statistic conceptually. I have used ft_freqstatistics in the past and would like to frame this bootstat statistic in a similar manner, which is why I'm asking here. Also, if I am misunderstanding my statistic, advice either on how to properly implement ft_freqstatistics-like cluster permutation statistics in this bootstat function, or alternatively how to convert my ersp matrix in such a way as to be usable with ft_freqstatistics, would be appreciated. Thanks, Max -- Max Cantor Graduate Student Cognitive Neuroscience of Language Lab University of Colorado Boulder -------------- next part -------------- An HTML attachment was scrubbed... URL: From o.jensen at bham.ac.uk Sat Oct 28 09:40:13 2017 From: o.jensen at bham.ac.uk (Ole Jensen) Date: Sat, 28 Oct 2017 08:40:13 +0100 Subject: [FieldTrip] 3 postdoc positions / U of Birmingham Message-ID: <7a17da5d-34c3-ad37-0a78-c67c68bfde5b@bham.ac.uk> Dear all, I have 3 postdoc openings funded by a Wellcome Trust Investigator Award. See here for more information. Applications can be submitted from Nov 1 to 16. ==== Applications are invited for a full-time Postdoctoral Research Fellow to work at the Centre for Human Brain Health, University of Birmingham. The successful candidate will work on a project aimed at studying the role of brain oscillations in relation to perception, attention and memory. This will be done in the group of Prof. Ole Jensen. The methodological focus of the project is MEG recordings and re-analysis of existing animal data. The candidate will be making use of the newly to be installed MEG system at the University of Birmingham and help to develop the laboratory. The candidate will be involved in all stages of designing and conducting experiments, analysing electrophysiological data. The candidate will be expected to prepare the results for high quality academic publications, to present at national and international academic conferences, and to engage in public activities. All research will be conducted in the spirit of open science. Furthermore the candidates will take part in supervision students and contribute to the Centre for Human Brain Health through leadership. === Feel free to write me for details. All the best, Ole -- Prof. Ole Jensen Centre for Human Brain Health, University of Birmingham http://www.neuosc.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From fereshte.ramezani at gmail.com Sun Oct 29 07:40:56 2017 From: fereshte.ramezani at gmail.com (Fereshte) Date: Sun, 29 Oct 2017 10:10:56 +0330 Subject: [FieldTrip] Import your own segmentation for making the head model in FiledTrip Message-ID: Dear Experts, How can one import his own segmentation data ( five labels as ; skull,scalp,GM,WM and CSF) into FiledTrip to make the finite element head model ? Thank you for your attention. Regards, Fereshte -------------- next part -------------- An HTML attachment was scrubbed... URL: From yoland.savriama at helsinki.fi Mon Oct 30 12:47:27 2017 From: yoland.savriama at helsinki.fi (Savriama, Yoland F) Date: Mon, 30 Oct 2017 11:47:27 +0000 Subject: [FieldTrip] How make a heat map on a skull using 3D landmark-based geometric morphometrics? Message-ID: Dear colleagues, I use 3D landmark-based geometric morphometrics for statistical analyses of morphological variation in vertebrate skulls. I would like to know how I could create a heat map of a skull that will show the deviations from the mean shape for a given effect around each landmark. For instance, the patterns of shape changes associated with a given Principal Component that would be mapped on a 3D skull representing the average mean shape. In this case, the visualisations of these deviations will be restricted and associated only with the local landmark coordinates as opposed to a full heat map that would cover the entire skull. In fact, the exact thing I would like to achieve had been done in Figure 4A of this publication: Maga, A. M., Navarro, N., Cunningham, M. L., & Cox, T. C. (2015). Quantitative trait loci affecting the 3D skull shape and size in mouse and prioritization of candidate genes in-silico. Frontiers in physiology, 6. https://doi.org/10.3389/fphys.2015.00092 I would greatly appreciate your feedback on this matter. Best wishes, Yoland SAVRIAMA, PhD Institute of Biotechnology P.O. Box 56 (Viikinkaari 5) FIN-00014 FINLAND -------------- next part -------------- An HTML attachment was scrubbed... URL: From mathilde.bonnefond at inserm.fr Mon Oct 30 17:07:04 2017 From: mathilde.bonnefond at inserm.fr (mathilde.bonnefond at inserm.fr) Date: Mon, 30 Oct 2017 17:07:04 +0100 Subject: [FieldTrip] postdoctoral position on the mechanistic role of brain oscillations In-Reply-To: <230d4c0ddd0a8ac1e161b1dbed847738@inserm.fr> References: <230d4c0ddd0a8ac1e161b1dbed847738@inserm.fr> Message-ID: Dear all, I'm offering a post-doc position on the role of nested oscillations in brain networks communication at the Lyon Neuroscience Research Center (CRNL) starting in February (see attached document). I would be grateful if you could share the attached job offer with interested PhD students and postdocs around you. Thank you very much. Best regards, Mathilde Bonnefond ---- Mathilde Bonnefond, PhD Lyon Neuroscience Research Center (CRNL) Inserm-CNRS-University Lyon 1 Dycog group Bron, France -------------- next part -------------- A non-text attachment was scrubbed... Name: PostDocAdvertisment.pdf Type: application/pdf Size: 232069 bytes Desc: not available URL: From heqing.psy at hotmail.com Tue Oct 31 08:45:06 2017 From: heqing.psy at hotmail.com (He Qing) Date: Tue, 31 Oct 2017 07:45:06 +0000 Subject: [FieldTrip] import Curry dataset into FieldTrip Message-ID: Dear FieldTripist: I am new to FiledTrip, could someone tell me how to read Curry dataset into FieldTrip? When I call the ft_preprocessing or ft_definetrial, it is alway showing errors. ---------------------------------------------- Error using ft_notification (line 340) unsupported header format "curry_dat" Qing He -------------- next part -------------- An HTML attachment was scrubbed... URL: From isac.sehlstedt at psy.gu.se Mon Oct 2 13:03:17 2017 From: isac.sehlstedt at psy.gu.se (Isac Sehlstedt) Date: Mon, 2 Oct 2017 11:03:17 +0000 Subject: [FieldTrip] Follow up question: Computing pca variables (i.e. Latent, and coefficient variables) after ft_componentanalysis (Schoffelen, J.M. (Jan Mathijs)) Message-ID: Dear Jan, I am such a blind hen sometimes. Not the proudest moment of my life. Thank you for helping me find what I was looking for. Very Best, Isac -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 09:33:12 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 07:33:12 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> Message-ID: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 09:39:04 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 07:39:04 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization Message-ID: <6d86574c600e4bc4a114b4d7ea788d18@neuro.gu.se> Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz -------------- next part -------------- An HTML attachment was scrubbed... URL: From sarang at cfin.au.dk Wed Oct 4 09:47:19 2017 From: sarang at cfin.au.dk (Sarang S. Dalal) Date: Wed, 4 Oct 2017 07:47:19 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> Message-ID: <1507103238.2609.8.camel@cfin.au.dk> Hi Bushra, Are you using the common filter approach to use the same spatial filter for your contrasted conditions? http://www.fieldtriptoolbox.org/example/common_filters_in_beamforming (Although that example is focused on DICS and PCC, a similar approach can also be used for LCMV.) This should reduce the likelihood of the 'middle of the head' artifact. If the filters are computed independently, however, then the noise suppression characteristics of each filter can be different, so the contrast would pull out differences in each filter's noise performance. If you are already using this approach, then you can try weight normalization, which is one strategy for removing the noise bias (and usually prevents the 'center of the head' artifact even when no contrast is done). You can run this by adding to your cfg structure: cfg.lcmv.weightnorm = 'nai' - or - cfg.lcmv.weightnorm = 'unitnoisegain'; Each of those methods should give you the same result with a contrast. The NAI is simply scaled against the theoretical noise power (i.e., the noise level = 1), so is often preferred when no contrast is done. Cheers, Sarang On Wed, 2017-10-04 at 07:33 +0000, Bushra Riaz Syeda wrote: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don’t see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From bushra.riaz2 at neuro.gu.se Wed Oct 4 10:19:09 2017 From: bushra.riaz2 at neuro.gu.se (Bushra Riaz Syeda) Date: Wed, 4 Oct 2017 08:19:09 +0000 Subject: [FieldTrip] Artifact in LCMV beamformer source localization In-Reply-To: <1507103238.2609.8.camel@cfin.au.dk> References: , <00543f8b8ead495087a4de0c95981a29@neuro.gu.se> , <1507103238.2609.8.camel@cfin.au.dk> Message-ID: Thank you for you reply I am already using common filter for both conditions and I have also tried weight normalisation with 'unitnoisegain', but it didn't change results. My protocol is bit unusual in a way that electric stimulation to the finger is time locked with 200 ms after the R wave of ECG. As such, the ECG artifact is prominent at -200 ms in my data and I am using ICA to remove it as shown in figure (https://chalmersuniversity.box.com/s/deuzucqovgcfienybk7zisc3r3wjusvu) For lcmv I am using -0.71 to -0.21 sec of prestimulus interval window ( relatively clean baseline without ECG artifact) and 0.01 to 0.51 sec of post stimulus interval window. Could my protocol be the reason of 'center of head artifact', I didnt think it would be the case? Any suggestions to solve this would be highly appreciated. Thank you Kind regards Bushra Riaz ________________________________ From: fieldtrip-bounces at science.ru.nl on behalf of Sarang S. Dalal Sent: Wednesday, October 4, 2017 9:47:19 AM To: fieldtrip at science.ru.nl Subject: Re: [FieldTrip] Artifact in LCMV beamformer source localization Hi Bushra, Are you using the common filter approach to use the same spatial filter for your contrasted conditions? http://www.fieldtriptoolbox.org/example/common_filters_in_beamforming (Although that example is focused on DICS and PCC, a similar approach can also be used for LCMV.) This should reduce the likelihood of the 'middle of the head' artifact. If the filters are computed independently, however, then the noise suppression characteristics of each filter can be different, so the contrast would pull out differences in each filter's noise performance. If you are already using this approach, then you can try weight normalization, which is one strategy for removing the noise bias (and usually prevents the 'center of the head' artifact even when no contrast is done). You can run this by adding to your cfg structure: cfg.lcmv.weightnorm = 'nai' - or - cfg.lcmv.weightnorm = 'unitnoisegain'; Each of those methods should give you the same result with a contrast. The NAI is simply scaled against the theoretical noise power (i.e., the noise level = 1), so is often preferred when no contrast is done. Cheers, Sarang On Wed, 2017-10-04 at 07:33 +0000, Bushra Riaz Syeda wrote: Hello I am using lcmv to source localize somatosensory response of Electric stimulation to the index finger of the left hand of subjects. 500 ms of post-stimulus interval is contrasted to 500 ms of pre-stimulus interval. The problem is I see strong activity in the middle of the head for couple of my subjects. It is very strong for one subject as shown ( link to figure 1: https://chalmersuniversity.box.com/s/jxmq5zp8k9m9tdfpogem7ee3r4z9q3ac )and week for other subjects ( link for figure 2: https://chalmersuniversity.box.com/s/o3yl57s6phr96f9rxh4wuthjcgllr2vv ) . I thought that noise bias towards the center of the head is usually taken care of when comparing conditions. Could somebody guide me how to solve this problem. Specially figure 1 where the noise bias is so strong that I don't see any activity in somatosensory cortex. Thank you Kind regards Bushra Riaz _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From na.so.ir at gmail.com Thu Oct 5 12:44:35 2017 From: na.so.ir at gmail.com (Narjes Soltani) Date: Thu, 5 Oct 2017 12:44:35 +0200 Subject: [FieldTrip] Median Filter Message-ID: Dear Fieldtripers, I have a question about the median filter. When I set this filter to 'no' (which is the default value) in ft_preprocessing configuration file and then run the ICA on the output, multiple components are not converged. However if I set the median filter to 'yes', this problem vanishes. Is there any idea why setting the median filter to 'yes' which means keeping the jump artifacts, helps in ICA convergence? By the way, I sent the epoched data to ft_preprocessing which may also be important about how ICA works. Very Best, Narjes -------------- next part -------------- An HTML attachment was scrubbed... URL: From chrichat at hotmail.com Fri Oct 6 15:42:02 2017 From: chrichat at hotmail.com (Christos Chatzichristos) Date: Fri, 6 Oct 2017 13:42:02 +0000 Subject: [FieldTrip] Simulation of sources based on fMRI spatial map. Message-ID: Dear all I am new at the community of field trip (and EEG ) generally. I am a PHD student and I am working on tensor representation for fMRI Blind Source Separation. Currently I am working on a parallel project for fusion of EEG and fMRI (with tensors again). Since there are new methods before going to real data I need some realistic simulated to test my methods on. I went through the tutorials about `Creating a sourcemodel for source-reconstruction of MEG or EEG data' and also the example 'Compute forward simulated data and apply a dipole fit' but still I was wondering if there is a way to create the leadfield matrices for sources simulated for fMRI or even more simply, from single slice sources like the one attached (which has been simulated with SimTB for a spherical brain). I was thinking to perform a simulation similar to this used in [X. Lei et al. 'A parallel framework for simultaneous EEG/fMRI analysis: Methodology and simulation], where they build a concentric three-sphere model around a spherical brain slice similar to the one I attach. Thanks a lot for any possible help. Kind regards Christos Chatzichristos -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Screenshot from 2017-10-06 14-37-38.png Type: image/png Size: 17053 bytes Desc: Screenshot from 2017-10-06 14-37-38.png URL: From aarjona at us.es Fri Oct 6 12:09:48 2017 From: aarjona at us.es (Antonio Arjona Valladares) Date: Fri, 06 Oct 2017 12:09:48 +0200 Subject: [FieldTrip] Fwd: the field cfg.neighbours is required In-Reply-To: References: Message-ID: Hello everyone! This is my query: Im trying to make statistical analyses of time-frequency with two groups of subjects. I have created the file .study with both groups, precompute de channels, and then I would like to apply the 'montecarlo/permutation statistics' with 'cluster correction'. The problem comes when I select all the electrodes and click on 'Plot ERSP(s)'. After a few seconds, the program gives me this answer (image 3). I have been reading tutorials, but nothings appears about this issue. Thank you so much --- Antonio Arjona Valladares Research Technician (BS) Human Psychobiology Lab, Experimental Psychology Department, University of Seville, C/Camilo José Cela s/n, 41018-Sevilla, Spain Tel: +34 954 55 78 00 - 954 55 69 41 -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image 3.png Type: image/png Size: 1067786 bytes Desc: not available URL: From abela.eugenio at gmail.com Fri Oct 6 17:30:50 2017 From: abela.eugenio at gmail.com (Eugenio Abela) Date: Fri, 6 Oct 2017 16:30:50 +0100 Subject: [FieldTrip] Fwd: the field cfg.neighbours is required In-Reply-To: References: Message-ID: <9A6240C6-2D77-4A26-BB76-3DDFD61CD417@gmail.com> Hi Antonio this looks like a question that EEGLAB might also be able to answer (https://sccn.ucsd.edu/wiki/EEGLAB_mailing_lists )? In any case, see here for what cfg.neighbours means /does: http://www.fieldtriptoolbox.org/tutorial/cluster_permutation_timelock http://www.fieldtriptoolbox.org/faq/how_can_i_define_neighbouring_sensors I’m sorry, but I don’t know how you specify the neighbour structure in EEGLAB... Good luck Eugenio On 6 Oct 2017, at 11:09, Antonio Arjona Valladares wrote: Hello everyone! This is my query: Im trying to make statistical analyses of time-frequency with two groups of subjects. I have created the file .study with both groups, precompute de channels, and then I would like to apply the 'montecarlo/permutation statistics' with 'cluster correction'. The problem comes when I select all the electrodes and click on 'Plot ERSP(s)'. After a few seconds, the program gives me this answer (image 3). I have been reading tutorials, but nothings appears about this issue. Thank you so much --- Antonio Arjona Valladares Research Technician (BS) Human Psychobiology Lab, Experimental Psychology Department, University of Seville, C/Camilo José Cela s/n, 41018-Sevilla, Spain Tel: +34 954 55 78 00 - 954 55 69 41 _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image 3.png Type: image/png Size: 231844 bytes Desc: not available URL: From cmhan616 at utexas.edu Tue Oct 10 21:54:57 2017 From: cmhan616 at utexas.edu (Chungmin Han) Date: Tue, 10 Oct 2017 14:54:57 -0500 Subject: [FieldTrip] Question on TFR analysis Message-ID: Dear all, I'm Chungmin, graduate student in UT Austin, and have a question TFR using fieldtrip, my experiment has 2 triggers, first one is Trial onset second trigger is the beginning of task with fixed time when subject trigger to begin task. Thus there's varying time between 1st and 2nd trigger, I need to analyze signal with fixed time window after 2nd trigger but need to take baseline right before 1st trigger because there's movement between 1st and 2nd, is there a way I can set up baseline window in matrix form?? seems TFRanalysis allows single baseline time window settings not for varying time window, I've also looked at redefine_trials, however, in this case, I'll have to put two inputs to calculate TFR, one as baselines only and the other task only. .....if either of the methods is not possible, can I set up single baseline (e.g. resting state data)??? Any advice is appreciated Thank you, Best Chungmin -------------- next part -------------- An HTML attachment was scrubbed... URL: From M.vanEs at donders.ru.nl Wed Oct 11 13:58:14 2017 From: M.vanEs at donders.ru.nl (Es, M.W.J. van (Mats)) Date: Wed, 11 Oct 2017 11:58:14 +0000 Subject: [FieldTrip] Question on TFR analysis (Chungmin Han) Message-ID: <3FC79061C73BEF44A3BEDA5DFC0ADBDF86529C4F@EXPRD99.hosting.ru.nl> Hi Chungmin, You are right, it is not possible to make your contrast in a single step. It is not possible to do a baseline correction in ft_freqbaseline (also holds for ft_singleplotTFR etc.) when the baseline is time locked to a different event than the 'active' condition. In order to get the contrast you should get the TFR for both 'active' and 'baseline' condition, timelocked to the different events (by using ft_redefinetrial). You can then use ft_math to calculate the difference. Good luck, Mats van Es -----Original Message----- From: fieldtrip-bounces at science.ru.nl [mailto:fieldtrip-bounces at science.ru.nl] On Behalf Of fieldtrip-request at science.ru.nl Sent: woensdag 11 oktober 2017 12:00 To: fieldtrip at science.ru.nl Subject: fieldtrip Digest, Vol 83, Issue 6 Send fieldtrip mailing list submissions to fieldtrip at science.ru.nl To subscribe or unsubscribe via the World Wide Web, visit https://mailman.science.ru.nl/mailman/listinfo/fieldtrip or, via email, send a message with subject or body 'help' to fieldtrip-request at science.ru.nl You can reach the person managing the list at fieldtrip-owner at science.ru.nl When replying, please edit your Subject line so it is more specific than "Re: Contents of fieldtrip digest..." Today's Topics: 1. Question on TFR analysis (Chungmin Han) ---------------------------------------------------------------------- Message: 1 Date: Tue, 10 Oct 2017 14:54:57 -0500 From: Chungmin Han To: fieldtrip at science.ru.nl Subject: [FieldTrip] Question on TFR analysis Message-ID: Content-Type: text/plain; charset="utf-8" Dear all, I'm Chungmin, graduate student in UT Austin, and have a question TFR using fieldtrip, my experiment has 2 triggers, first one is Trial onset second trigger is the beginning of task with fixed time when subject trigger to begin task. Thus there's varying time between 1st and 2nd trigger, I need to analyze signal with fixed time window after 2nd trigger but need to take baseline right before 1st trigger because there's movement between 1st and 2nd, is there a way I can set up baseline window in matrix form?? seems TFRanalysis allows single baseline time window settings not for varying time window, I've also looked at redefine_trials, however, in this case, I'll have to put two inputs to calculate TFR, one as baselines only and the other task only. .....if either of the methods is not possible, can I set up single baseline (e.g. resting state data)??? Any advice is appreciated Thank you, Best Chungmin -------------- next part -------------- An HTML attachment was scrubbed... URL: ------------------------------ _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip End of fieldtrip Digest, Vol 83, Issue 6 **************************************** From luca.turella at unitn.it Wed Oct 11 17:06:13 2017 From: luca.turella at unitn.it (Luca Turella) Date: Wed, 11 Oct 2017 17:06:13 +0200 Subject: [FieldTrip] Post-doc in Neural dynamics of Action understanding @ CIMeC University of Trento, Italy Message-ID: A postdoctoral position will be available soon at the Center for Mind/Brain Sciences (CIMeC, http://www.cimec.unitn.it/en) at the University of Trento (Italy). The topic of investigation will cover the neural dynamics underlying action understanding adopting behavioural studies and MEG. The requirements for the candidates are the following. Candidates must have a Ph.D. degree in a field related to Cognitive Neuroscience or related areas. Previous experience in behavioural/kinematic studies and/or EEG/MEG data analysis is required. The ideal candidate should have good programming skills in Matlab, preferably related to EEG/MEG data analysis (Fieldtrip). Knowledge of kinematic data analysis is also a plus. Knowledge of Italian language is not required. The salary will be proportional to the level of experience and the starting date of the appointment is negotiable, but within the next 6 months. Applications will be considered until the position is filled. The contract will have a duration of 2 years. Applications should be sent to luca.turella at unitn.it, including complete CV, statement of research interests, and contact details of two referees. Potential candidates are also encouraged to send me directly informal inquires *luca.turella at unitn.it *. CIMeC offers an international and vibrant research setting with access to state-of-the-art neuroimaging methodologies, including a research-only MR scanner, MEG, EEG and TMS, as well as behavioural, eye tracking and motion tracking laboratories. English is the official language of the CIMeC, where a large proportion of the faculty, post-docs and students come from a wide range of countries outside of Italy. The University of Trento consistently ranks as a top Italian university in both national Research Assessment Evaluations (RAE) and University Surveys. In the latest RAE, the University of Trento as a whole ranks 2nd among medium-sized universities. -- Luca Turella, PhD Assistant Professor CIMeC - Center for Mind/Brain Sciences University of Trento Mattarello (TN), Via Delle Regole 101 Tel.+39 0461-28 3098 http://www.unitn.it/cimec Legal Disclaimer This electronic message contains information that is confidential. The information is intended for the use of the addressee only. If you are not the addressee we would appreciate your notification in this respect. Please note that any disclosure, copy, distribution or use of the contents of this message is prohibited and may be unlawful. Avvertenza legale Questo messaggio Email contiene informazioni confidenziali riservate ai soli destinatari. Qualora veniate in possesso di tali informazioni senza essere definito come destinatario vi reghiamo di leggere le seguenti note. Ogni apertura, copia, distribuzione del contenuto del messaggio e dei suoi allegati è proibito e potrebbe violare le presenti leggi. -------------- next part -------------- An HTML attachment was scrubbed... URL: From m.stoica at uke.de Wed Oct 11 21:54:41 2017 From: m.stoica at uke.de (Mircea Stoica) Date: Wed, 11 Oct 2017 21:54:41 +0200 Subject: [FieldTrip] Sharp transition in multitaper coherence when moving from one taper to two Message-ID: Dear fieldtrippers, I'm interested in coherence analysis with mtmconvol and multitapers. My frequencies of interest are spaced logarithmically (steps of 1/16 octaves) and I adjust the time window and number of tapers to yield a constant frequency smoothing of 0.75 octaves (half that for cfg.tapsmofrq of course). This results in an ugly step-like transition of coherence when moving from one taper to two, as you can see in the following image. https://photos.app.goo.gl/5uWy05RzZxaA9gci1 This sharp transition is not visible in the power output. I calculated the actual frequency smoothing as fw = (K + 1) ./ cfg.t_ftimwin / 2 and in a semilogy plot it gives this: https://photos.app.goo.gl/xQsrY1jr8L0qLLDm2 And the number of tapers for each frequency: https://photos.app.goo.gl/LqaeUVW6ymBWEKEx2 There are no discontinuities in the frequency smoothing. Any ideas what could be the cause? Or is there perhaps nothing to be done about it? Best, Mircea -- _____________________________________________________________________ Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg | www.uke.de Vorstandsmitglieder: Prof. Dr. Burkhard Göke (Vorsitzender), Prof. Dr. Dr. Uwe Koch-Gromus, Joachim Prölß, Martina Saurin (komm.) _____________________________________________________________________ SAVE PAPER - THINK BEFORE PRINTING -------------- next part -------------- An HTML attachment was scrubbed... URL: From stephen.whitmarsh at gmail.com Fri Oct 13 17:29:35 2017 From: stephen.whitmarsh at gmail.com (Stephen Whitmarsh) Date: Fri, 13 Oct 2017 17:29:35 +0200 Subject: [FieldTrip] NAI on trial-by-trail power estimates Message-ID: Dear all, I am trying to optimize my beamformer power estimates, which are not bad, but I want to see if I can improve them. I do not seem to have a depth-bias, and am testing within-subject between-condition contrasts, for which one does not need to correct for noise differences. However, my analysis does involve trial-by-trial analysis, and might be susceptible to noise differences over time. I was wondering whether it would make sense to have trial-by-trial corrections using the Neural Activation Index (NAI), as described in the beamformer tutorial, i.e. by dividing my single-trial power estimates by single-trial noise estimates. Has anyone tried using either the NAI on a trial-by-trial basis? Secondly, does this even make sense to you? Cheers, Stephen -------------- next part -------------- An HTML attachment was scrubbed... URL: From yuxxx955 at umn.edu Mon Oct 16 23:26:30 2017 From: yuxxx955 at umn.edu (Kai Yu) Date: Mon, 16 Oct 2017 16:26:30 -0500 Subject: [FieldTrip] How to understand the cross-correlogram and its feature comparison against the one obtained by shuffling the trials Message-ID: Hi, I have been using your FieldTrip Intracranial Spike Analysis package for a while. However, I am still confused by the statement listed on your website. "Cross-correlations between neurons can either arise because of common, time-locked fluctuations in the firing rate (Brody et al., 1999). These correlations are invariant to a change in the order of trials. The shuffling of trials in ft_spike_xcorr always pertains to two subsequent trials, in order to avoid an influence of slow changes in the firing rate across trials. We refer to this cross-correlogram that is obtained under a permutation of subsequent trials as the 'shift-predictor' cross-correlogram. If the observed features of the cross-correlogram that are not present in the shift-predictor cross-correlogram, then this indicates that they arise because of induced synchronous activity." For the last sentence, would you please explained more about the "induced synchronous activity"? In my own understanding, this is the activity after the task or external stimulation. But if I want to know the synchronous or de-synchronous activity during the stimulation, how can I get the evidence from these cross-correlations and also the jpsth figures? Thank you so much! Best, -Kai -- Kai Yu Biomedical Functional Imaging and Neuroengineering Laboratory Department of Biomedical Engineering University of Minnesota - Twin Cities -------------- next part -------------- An HTML attachment was scrubbed... URL: From acskwara at ucdavis.edu Wed Oct 18 23:05:45 2017 From: acskwara at ucdavis.edu (Alea Skwara) Date: Wed, 18 Oct 2017 14:05:45 -0700 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) Message-ID: Hi All, I have been conducting a permutation based cluster analysis using the method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to examine longitudinal changes in spectral power. Everything runs just fine and it identifies clusters successfully. However, when I examine the results of the analysis, some of the stats strike me as strange. Electrodes that have the highest F values (I am comparing across 3 timepoints) do not test as significant, while electrodes with lower F values do. At first I thought this was because I had a single "hot" electrode that was not part of a larger cluster, but looking more closely at the stats, it appears that this is not the case. I have tried double-checking my neighbor definition method and electrode locations to assure that it is not an issue with neighbors being incorrectly registered. I have also tried adjusting the minimum cluster size using the minnbchan option to ensure it was not an issue of a cluster not reaching the minimum extent. Neither of these options seem to have any impact on the result. At this point I'm at a loss. Is there something I'm misunderstanding about the algorithm? Code below and plot below, channel-by-channel stats attached. Thanks! Alea One funky thing to note about this code is that I am inputting average values into the analysis (averaged across time as participants have slightly differing lengths of data, and within bands as I am running this analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band power", not 7 Hz) %% THREE CONDITIONS %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) cfg = []; cfg.channel = {'all'}; cfg.latency = 'all'; cfg.frequency = [7 7]; %frequency band indicator cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing to average cfg.method = 'montecarlo'; cfg.statistic = 'depsamplesFmultivariate'; %three time points cfg.correctm = 'cluster'; cfg.clusteralpha = 0.1; %change to .1 because we're computing over averaged bands and time (losing power) cfg.clusterstatistic = 'maxsum'; cfg.minnbchan = 2; cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 otherwise cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account for 500 tests and correction by Bonferroni method. %specify which electrodes can form clusters cfg_neighb.method = 'distance'; cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); %cfg.neighbours = cfgManish.neighbours; %cfg = getNeighbours(cfg); % using fixed neighbours cfg.layout = layout'; % Set up the design matrixfor 3 conditions subj = 27; %will differ based on condition design = zeros(2,3*subj); for isub = 1:1:subj design(1,isub) = isub; design(1,subj+isub) = isub; design(1,2*subj+isub) = isub; end design(2,1:subj) = 1; design(2,subj+1:2*subj) = 2; design(2,2*subj+1:3*subj) = 3; cfg.design = design; cfg.uvar = 1; cfg.ivar = 2; % Run the analysis [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, r1t3r_struct); %Plot the results cfg1 = []; cfg1.alpha = 0.1; cfg1.parameter = 'stat'; %cfg1.zlim = [0 18.1]; cfg1.layout = layout'; cfg1.marker = 'o'; cfg1.subplotsize = [1 1]; cfg1.saveaspng = 'r1r_betacluster'; ft_clusterplot(cfg1, r1r_stat_beta); ***Note non-significant hot spot over left parietal*** [image: Inline image 2] Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: r1r_betacluster_fig1.png Type: image/png Size: 49573 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: clusterstats_beta.xlsx Type: application/vnd.openxmlformats-officedocument.spreadsheetml.sheet Size: 11196 bytes Desc: not available URL: From bioeng.yoosofzadeh at gmail.com Thu Oct 19 05:34:09 2017 From: bioeng.yoosofzadeh at gmail.com (Vahab Yousofzadeh) Date: Wed, 18 Oct 2017 23:34:09 -0400 Subject: [FieldTrip] Postdoctoral position at Ulster on Dementia and neuroimaging Message-ID: Hi everyone, There is a postdoc position at Ulster Univerity with Dr. KongFatt Wong-Lin and Prof. Tony Bjourson on dementia and neuroimaging. Check it out: https://goo.gl/4YexD5 Cheers, Vahab From jan.schoffelen at donders.ru.nl Thu Oct 19 08:19:35 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Thu, 19 Oct 2017 06:19:35 +0000 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) In-Reply-To: References: Message-ID: Hi Alea, I think that it all looks OK. I guess that the P5-PO1-O1 cluster is spatially disjoint from the bulk of the suprathreshold data points. Don’t be mislead by the topographical image, where both the contour lines, and the 2D placement of the electrodes might reflect a picture that is different from what you would expect. Note, that providing cirange as extra information is not the most informative. Better would have been to show posclusterslabelmat. This data field provides for each sample, whether and to which ‘cluster’ it belongs. Best wishes, Jan-Mathijs J.M.Schoffelen, MD PhD Senior Researcher, VIDI-fellow - PI, language in interaction Telephone: +31-24-3614793 Physical location: room 00.028 Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands On 18 Oct 2017, at 23:05, Alea Skwara > wrote: Hi All, I have been conducting a permutation based cluster analysis using the method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to examine longitudinal changes in spectral power. Everything runs just fine and it identifies clusters successfully. However, when I examine the results of the analysis, some of the stats strike me as strange. Electrodes that have the highest F values (I am comparing across 3 timepoints) do not test as significant, while electrodes with lower F values do. At first I thought this was because I had a single "hot" electrode that was not part of a larger cluster, but looking more closely at the stats, it appears that this is not the case. I have tried double-checking my neighbor definition method and electrode locations to assure that it is not an issue with neighbors being incorrectly registered. I have also tried adjusting the minimum cluster size using the minnbchan option to ensure it was not an issue of a cluster not reaching the minimum extent. Neither of these options seem to have any impact on the result. At this point I'm at a loss. Is there something I'm misunderstanding about the algorithm? Code below and plot below, channel-by-channel stats attached. Thanks! Alea One funky thing to note about this code is that I am inputting average values into the analysis (averaged across time as participants have slightly differing lengths of data, and within bands as I am running this analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band power", not 7 Hz) %% THREE CONDITIONS %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) cfg = []; cfg.channel = {'all'}; cfg.latency = 'all'; cfg.frequency = [7 7]; %frequency band indicator cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing to average cfg.method = 'montecarlo'; cfg.statistic = 'depsamplesFmultivariate'; %three time points cfg.correctm = 'cluster'; cfg.clusteralpha = 0.1; %change to .1 because we're computing over averaged bands and time (losing power) cfg.clusterstatistic = 'maxsum'; cfg.minnbchan = 2; cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 otherwise cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account for 500 tests and correction by Bonferroni method. %specify which electrodes can form clusters cfg_neighb.method = 'distance'; cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); %cfg.neighbours = cfgManish.neighbours; %cfg = getNeighbours(cfg); % using fixed neighbours cfg.layout = layout'; % Set up the design matrixfor 3 conditions subj = 27; %will differ based on condition design = zeros(2,3*subj); for isub = 1:1:subj design(1,isub) = isub; design(1,subj+isub) = isub; design(1,2*subj+isub) = isub; end design(2,1:subj) = 1; design(2,subj+1:2*subj) = 2; design(2,2*subj+1:3*subj) = 3; cfg.design = design; cfg.uvar = 1; cfg.ivar = 2; % Run the analysis [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, r1t3r_struct); %Plot the results cfg1 = []; cfg1.alpha = 0.1; cfg1.parameter = 'stat'; %cfg1.zlim = [0 18.1]; cfg1.layout = layout'; cfg1.marker = 'o'; cfg1.subplotsize = [1 1]; cfg1.saveaspng = 'r1r_betacluster'; ft_clusterplot(cfg1, r1r_stat_beta); ***Note non-significant hot spot over left parietal*** Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From maximilien.chaumon at gmail.com Thu Oct 19 11:18:51 2017 From: maximilien.chaumon at gmail.com (Maximilien Chaumon) Date: Thu, 19 Oct 2017 09:18:51 +0000 Subject: [FieldTrip] memory leak in ft_selectdata Message-ID: Dear all, I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion cfg = []; cfg.parameter = 'logmpowspctrm'; cfg.comment = 'no'; cfg.xlim = [20 21]; cfg.zlim = 'maxmin'; cfg.layout = 'neuromag306all.lay'; cfg.colorbar = 'no'; big_spec_grad.dimord = 'chan_freq'; ft_topoplotER(cfg,big_spec_grad); This used to work fine until a last update this week from a version I suspect dated back to 03/17. Now today, the ft_topoplotER line eats up a lot of memory the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. x = x(:,selindx,:,:,:,:); The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... Any help is much appreciated. Thanks! Max -------------- next part -------------- An HTML attachment was scrubbed... URL: From adonay.s.nunes at gmail.com Thu Oct 19 23:41:53 2017 From: adonay.s.nunes at gmail.com (A Nunes) Date: Thu, 19 Oct 2017 14:41:53 -0700 Subject: [FieldTrip] SAM gareth method without depth biased weights Message-ID: Hi community, I have an issue with Fieltrip SAM beamformer 'gareth' method. As a sanity check I look that the BF weight distributions are biased towards the center. I check this distribution with the leadfields and with the BF weights from LCMV and SAMs methods. Gareth method does not have this distribution, the weight sizes are scattered across the brain and the norm of the weights can have up to 11 orders of magnitude of difference. First I make sure that the 1/Leadfield are also biased towards the center: - for every LF I take the norm of every column and 1/mean of the norms is the value. - the code is: LF_norm = []; for s= 1:size(LF, 1) n1 = norm(squeeze(LF(s, 1, :))); n2 = norm(squeeze(LF(s, 2, :))); n3 = norm(squeeze(LF(s, 3, :))); LF_norm(s) =1/mean([n1, n2, n3]); end When interpolated to the mri and plotted looks fine. Then I run SAM robert method: cfg = []; cfg.method = 'sam'; cfg.grid = lf; cfg.vol = headmodel; cfg.grad = data.grad; cfg.fixedori = 'robert'; cfg.keepfilter = 'yes'; cfg.keepori = 'yes'; cfg.projectnoise = 'yes'; cfg.senstype = 'MEG'; SAM1_filter = ft_sourceanalysis(cfg, timelock); and take the norm for every BF weight: SAM1_filter_norm= cellfun(@(X) norm(X),SAM1_filter.avg.filter(SAM1_filter.inside)); and then I interpolate it to the mri for visualization. The results look fine, weights biased towards the center and the min and max norm weight are two orders of magnitude difference. The mean of the absolute of the weights norm is in the order of 10e9 and the smallest and highest weight norm are to the order of e8 and e10 If I do the same as the last step but changing to: cfg.fixedori = 'gareth'; Then the weights are not biased towards the center. The mean of the absolute of the weights norm is in the order of 10e16 and the smallest and highest weight norm are to the order e8 and e19, respectively. Thus, gareth method does not have a center biased weight distribution and the max and miminum norm weights have 11 orders of magnitude of difference, meaning that these orders are reflected in the amplitude of the reconstructed time series. To make it handy the difference between gareth and robert methods are: L is the leadfield for a given source. 'Gareth' Y1 = L' * inv_cov * L; Y2 = L' * (inv_cov * inv_cov) * L; [U,S] = eig(Y2,Y1); 'Robert' [U,S] = svd(real(pinv(L' * inv_cov * L))); Then both methods compute 2 of the orientations and choose the salient one: ori1 = U(:,1); ori1 = ori1/norm(ori1); ori2 = U(:,2); ori2 = ori2/norm(ori2); L1 = L * ori1; L2 = L * ori2; if (norm(L1)/norm(L2)) < 1e-6 opt_vox_or = ori2; else opt_vox_or = ori1; end The difference comes from the way orientations are computed, after the weights calculation is the same for both. I tried with different units, i.e. meters, cm, mm but this abnormalities is still present. Does anybody know what can be the cause? Thanks Adonay -------------- next part -------------- An HTML attachment was scrubbed... URL: From jan.schoffelen at donders.ru.nl Fri Oct 20 09:07:10 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Fri, 20 Oct 2017 07:07:10 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: Hi Max, >From your (otherwise clear) description, I am not sure whether your issue is due to the actual data selection itself. Rather, my first thought would be that something goes amiss with recursive piling up of large amounts of irrelevant (for the plotting at least) metadata, that gets assembled while iterating through subsequent interactive plotting steps. It’s kind of a long story, which I am not going to type down here, but to check whether I am thinking into the right direction, could you see whether your problem persists if you call ft_topoplotER in the following way? ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); Thanks, Jan-Mathijs > On 19 Oct 2017, at 11:18, Maximilien Chaumon wrote: > > Dear all, > > I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion > cfg = []; > cfg.parameter = 'logmpowspctrm'; > cfg.comment = 'no'; > cfg.xlim = [20 21]; > cfg.zlim = 'maxmin'; > cfg.layout = 'neuromag306all.lay'; > cfg.colorbar = 'no'; > big_spec_grad.dimord = 'chan_freq'; > ft_topoplotER(cfg,big_spec_grad); > > This used to work fine until a last update this week from a version I suspect dated back to 03/17. > Now today, the ft_topoplotER line eats up a lot of memory > the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB > the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB > the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB > > This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. > > I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. > x = x(:,selindx,:,:,:,:); > The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). > Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... > > Any help is much appreciated. > Thanks! > Max > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip From mibanks at wisc.edu Fri Oct 20 16:16:54 2017 From: mibanks at wisc.edu (MATTHEW I BANKS) Date: Fri, 20 Oct 2017 14:16:54 +0000 Subject: [FieldTrip] Variance estimates for connectivity measures Message-ID: Greetings. Is it possible to use jackknife to get variance estimates for Granger, partial directed coherence and directed transfer function? I use it for wPLI, but cannot figure out how to do it for these other measures. -Matt Banks ____________________________ Matthew I. Banks, Ph.D. Associate Professor Department of Anesthesiology University of Wisconsin 1300 University Avenue, Room 4605 Madison, WI 53706 office tel. (608)261-1143 lab tel. (608)263-6662 fax (608)263-2592 http://anesthesia.wisc.edu/index.php/Banks_Laboratory http://ntp.neuroscience.wisc.edu/banks.htm -------------- next part -------------- An HTML attachment was scrubbed... URL: From maximilien.chaumon at gmail.com Fri Oct 20 16:30:19 2017 From: maximilien.chaumon at gmail.com (Maximilien Chaumon) Date: Fri, 20 Oct 2017 14:30:19 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: Perfect! that fixed it, many thanks! Max Le ven. 20 oct. 2017 à 09:35, Schoffelen, J.M. (Jan Mathijs) < jan.schoffelen at donders.ru.nl> a écrit : > Hi Max, > > From your (otherwise clear) description, I am not sure whether your issue > is due to the actual data selection itself. Rather, my first thought would > be that something goes amiss with recursive piling up of large amounts of > irrelevant (for the plotting at least) metadata, that gets assembled while > iterating through subsequent interactive plotting steps. > It’s kind of a long story, which I am not going to type down here, but to > check whether I am thinking into the right direction, could you see whether > your problem persists if you call ft_topoplotER in the following way? > > > ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); > > Thanks, > > Jan-Mathijs > > > > > On 19 Oct 2017, at 11:18, Maximilien Chaumon < > maximilien.chaumon at gmail.com> wrote: > > > > Dear all, > > > > I'm doing topoplots of rather large spectrum data focused on a tiny > frequency portion > > cfg = []; > > cfg.parameter = 'logmpowspctrm'; > > cfg.comment = 'no'; > > cfg.xlim = [20 21]; > > cfg.zlim = 'maxmin'; > > cfg.layout = 'neuromag306all.lay'; > > cfg.colorbar = 'no'; > > big_spec_grad.dimord = 'chan_freq'; > > ft_topoplotER(cfg,big_spec_grad); > > > > This used to work fine until a last update this week from a version I > suspect dated back to 03/17. > > Now today, the ft_topoplotER line eats up a lot of memory > > the call to "ft_prepare_layout" took 0 seconds and required the > additional allocation of an estimated 0 MB > > the call to "ft_selectdata" took 0 seconds and required the additional > allocation of an estimated 1283 MB > > the call to "ft_topoplotER" took 1 seconds and required the additional > allocation of an estimated 1283 MB > > > > This gets more dramatic the second time this part of the script runs > (with 2500MB added for reach topoplot) and then crashes on the next > iteration with all of 30GB RAM eaten up by this. > > > > I traced down the memory leak to ft_selectdata line 1281 in a > subfunction called cellmatselect. > > x = x(:,selindx,:,:,:,:); > > The memory usage jumps every time this dimension gets selected. (I admit > I do have a lot of frequencies, over 13000, but I need them for my > application). > > Memory comes back to a normal level when the figure is closed, but as I > said, execution with another dataset eats up double the amount of memory > and I don't know how to get out of this... > > > > Any help is much appreciated. > > Thanks! > > Max > > _______________________________________________ > > fieldtrip mailing list > > fieldtrip at donders.ru.nl > > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From jan.schoffelen at donders.ru.nl Fri Oct 20 16:40:34 2017 From: jan.schoffelen at donders.ru.nl (Schoffelen, J.M. (Jan Mathijs)) Date: Fri, 20 Oct 2017 14:40:34 +0000 Subject: [FieldTrip] memory leak in ft_selectdata In-Reply-To: References: Message-ID: <68BED22E-0754-4991-B4D5-4F9EAE8536CE@donders.ru.nl> You’re welcome. Yet, I wouldn’t say it’s fixed, but at least you have found a way around your problem. Best wishes, Jan-Mathijs On 20 Oct 2017, at 16:30, Maximilien Chaumon > wrote: Perfect! that fixed it, many thanks! Max Le ven. 20 oct. 2017 à 09:35, Schoffelen, J.M. (Jan Mathijs) > a écrit : Hi Max, From your (otherwise clear) description, I am not sure whether your issue is due to the actual data selection itself. Rather, my first thought would be that something goes amiss with recursive piling up of large amounts of irrelevant (for the plotting at least) metadata, that gets assembled while iterating through subsequent interactive plotting steps. It’s kind of a long story, which I am not going to type down here, but to check whether I am thinking into the right direction, could you see whether your problem persists if you call ft_topoplotER in the following way? ft_topoplotER(cfg, rmfield(big_spec_grad, ‘cfg’)); Thanks, Jan-Mathijs > On 19 Oct 2017, at 11:18, Maximilien Chaumon > wrote: > > Dear all, > > I'm doing topoplots of rather large spectrum data focused on a tiny frequency portion > cfg = []; > cfg.parameter = 'logmpowspctrm'; > cfg.comment = 'no'; > cfg.xlim = [20 21]; > cfg.zlim = 'maxmin'; > cfg.layout = 'neuromag306all.lay'; > cfg.colorbar = 'no'; > big_spec_grad.dimord = 'chan_freq'; > ft_topoplotER(cfg,big_spec_grad); > > This used to work fine until a last update this week from a version I suspect dated back to 03/17. > Now today, the ft_topoplotER line eats up a lot of memory > the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB > the call to "ft_selectdata" took 0 seconds and required the additional allocation of an estimated 1283 MB > the call to "ft_topoplotER" took 1 seconds and required the additional allocation of an estimated 1283 MB > > This gets more dramatic the second time this part of the script runs (with 2500MB added for reach topoplot) and then crashes on the next iteration with all of 30GB RAM eaten up by this. > > I traced down the memory leak to ft_selectdata line 1281 in a subfunction called cellmatselect. > x = x(:,selindx,:,:,:,:); > The memory usage jumps every time this dimension gets selected. (I admit I do have a lot of frequencies, over 13000, but I need them for my application). > Memory comes back to a normal level when the figure is closed, but as I said, execution with another dataset eats up double the amount of memory and I don't know how to get out of this... > > Any help is much appreciated. > Thanks! > Max > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl https://mailman.science.ru.nl/mailman/listinfo/fieldtrip -------------- next part -------------- An HTML attachment was scrubbed... URL: From acskwara at ucdavis.edu Fri Oct 20 20:59:49 2017 From: acskwara at ucdavis.edu (Alea Skwara) Date: Fri, 20 Oct 2017 11:59:49 -0700 Subject: [FieldTrip] Problems with permutation based cluster analysis (ft_freqstatistics) In-Reply-To: References: Message-ID: Hi J.M., Thanks so much for your reply. The tip to not trust the topographic maps too much helped in my sleuthing. I realized that adjusting the minnbchan parameter actually *does* result in the topographic changes I thought I should see if I was understanding the algorithm correctly (and that I just wasn't seeing these changes in the map that I was using as a quick-check, but that they were present in the stats themselves). Problem solved! Thank you! Alea On Wed, Oct 18, 2017 at 11:19 PM, Schoffelen, J.M. (Jan Mathijs) < jan.schoffelen at donders.ru.nl> wrote: > Hi Alea, > > I think that it all looks OK. I guess that the P5-PO1-O1 cluster is > spatially disjoint from the bulk of the suprathreshold data points. Don’t > be mislead by the topographical image, where both the contour lines, and > the 2D placement of the electrodes might reflect a picture that is > different from what you would expect. > > Note, that providing cirange as extra information is not the most > informative. Better would have been to show posclusterslabelmat. This data > field provides for each sample, whether and to which ‘cluster’ it belongs. > > Best wishes, > Jan-Mathijs > > > J.M.Schoffelen, MD PhD > Senior Researcher, VIDI-fellow - PI, language in interaction > Telephone: +31-24-3614793 <+31%2024%20361%204793> > Physical location: room 00.028 > Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands > > > > On 18 Oct 2017, at 23:05, Alea Skwara wrote: > > Hi All, > > I have been conducting a permutation based cluster analysis using the > method = 'montecarlo' correctm= 'cluster' options in ft_freqstatistics to > examine longitudinal changes in spectral power. Everything runs just fine > and it identifies clusters successfully. > > However, when I examine the results of the analysis, some of the stats > strike me as strange. Electrodes that have the highest F values (I am > comparing across 3 timepoints) do not test as significant, while electrodes > with lower F values do. At first I thought this was because I had a single > "hot" electrode that was not part of a larger cluster, but looking more > closely at the stats, it appears that this is not the case. > > I have tried double-checking my neighbor definition method and electrode > locations to assure that it is not an issue with neighbors being > incorrectly registered. I have also tried adjusting the minimum cluster > size using the minnbchan option to ensure it was not an issue of a cluster > not reaching the minimum extent. Neither of these options seem to have any > impact on the result. > > At this point I'm at a loss. Is there something I'm misunderstanding about > the algorithm? > > Code below and plot below, channel-by-channel stats attached. > > Thanks! > Alea > > One funky thing to note about this code is that I am inputting average > values into the analysis (averaged across time as participants have > slightly differing lengths of data, and within bands as I am running this > analysis on IAF-based bands, thus "7" is a dummy code for "mean beta band > power", not 7 Hz) > > > %% THREE CONDITIONS > %Set up the analysis cfg for 3 conditions (pre- to mid- to post-) > cfg = []; > cfg.channel = {'all'}; > cfg.latency = 'all'; > cfg.frequency = [7 7]; %frequency band indicator > cfg.avgoverfreq = 'no'; %working on bands not frequencies, so nothing > to average > cfg.method = 'montecarlo'; > cfg.statistic = 'depsamplesFmultivariate'; %three time points > cfg.correctm = 'cluster'; > cfg.clusteralpha = 0.1; %change to .1 because we're computing over > averaged bands and time (losing power) > cfg.clusterstatistic = 'maxsum'; > cfg.minnbchan = 2; > cfg.tail = 1; %this has to be 1 for depsamplesF, 0 otherwise > cfg.clustertail = 1; %this has to be 1 for depsamplesF, 0 otherwise > cfg.alpha = 0.05; %this has to be 0.05 for depsamplesF, 0.025 > otherwise > cfg.numrandomization = 10000; %to get a min p-value of 0.0001 to account > for 500 tests and correction by Bonferroni method. > %specify which electrodes can form clusters > cfg_neighb.method = 'distance'; > cfg.neighbours = ft_prepare_neighbours(cfg_neighb, r1t1r_struct); > %cfg.neighbours = cfgManish.neighbours; > %cfg = getNeighbours(cfg); % using fixed neighbours > cfg.layout = layout'; > > > % Set up the design matrixfor 3 conditions > subj = 27; %will differ based on condition > > design = zeros(2,3*subj); > for isub = 1:1:subj > design(1,isub) = isub; > design(1,subj+isub) = isub; > design(1,2*subj+isub) = isub; > end > design(2,1:subj) = 1; > design(2,subj+1:2*subj) = 2; > design(2,2*subj+1:3*subj) = 3; > > cfg.design = design; > cfg.uvar = 1; > cfg.ivar = 2; > > > % Run the analysis > [r1r_stat_beta] = ft_freqstatistics(cfg, r1t1r_struct, r1t2r_struct, > r1t3r_struct); > > > %Plot the results > cfg1 = []; > cfg1.alpha = 0.1; > cfg1.parameter = 'stat'; > %cfg1.zlim = [0 18.1]; > cfg1.layout = layout'; > cfg1.marker = 'o'; > cfg1.subplotsize = [1 1]; > cfg1.saveaspng = 'r1r_betacluster'; > > ft_clusterplot(cfg1, r1r_stat_beta); > > > ***Note non-significant hot spot over left parietal*** > > > > > > > Alea C. Skwara > Graduate Student | Saron Lab > http://mindbrain.ucdavis.edu/labs/Saron > > Center for Mind and Brain | University of California, Davis > 267 Cousteau Place | Davis CA 95616 > Cell: (828) 273-8595 > > > > > > > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > > > > > > > _______________________________________________ > fieldtrip mailing list > fieldtrip at donders.ru.nl > https://mailman.science.ru.nl/mailman/listinfo/fieldtrip > -- Alea C. Skwara Graduate Student | Saron Lab http://mindbrain.ucdavis.edu/labs/Saron Center for Mind and Brain | University of California, Davis 267 Cousteau Place | Davis CA 95616 Cell: (828) 273-8595 -------------- next part -------------- An HTML attachment was scrubbed... URL: From caschera at dis.uniroma1.it Tue Oct 24 13:26:42 2017 From: caschera at dis.uniroma1.it (Stefano Caschera) Date: Tue, 24 Oct 2017 13:26:42 +0200 Subject: [FieldTrip] Problem with OpenMEEG Message-ID: Hi, I am using Matlab 2012b on Windows7 64bit. I have installed OpenMEEG.exe but I had this warning: *Warning! PATH too long installer unable to modify PATH!* So I tried through OpenMEEG.tar.gz, but when I try to follow the tutorial I have this warning: *'om_assemble' is not recognized as an internal or external command, * *operable program or batch file.* I added all OpenMeeg folder with subfolders to the Matlab search path, but If I write *system ('om_assemble') *the answer is the expected one only when I am inside the folder /bin. Any suggestions? Thanks in advance. Thank you, Stefano -- Stefano Caschera, PhD Student *Neuroelectrical Imaging and BCI lab* Fondazione Santa Lucia, IRCCS Via Ardeatina, 306 I-00179, Rome, Italy *Department of Computer, Control, andManagement Engineering "Antonio Ruberti"* Sapienza, University of Rome V. Ariosto, 25 00185, Rome, Italy Tel +39 06 5150 1165 Email: stefano.caschera at uniroma1.it caschera at dis.uniroma1.it s.caschera at hsantalucia.it ___________________________________________________ Mail priva di virus. www.avg.com <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> -------------- next part -------------- An HTML attachment was scrubbed... URL: From caschera at dis.uniroma1.it Tue Oct 24 14:02:33 2017 From: caschera at dis.uniroma1.it (Stefano Caschera) Date: Tue, 24 Oct 2017 14:02:33 +0200 Subject: [FieldTrip] Problem with OpenMEEG Message-ID: Hi, I am using Matlab 2012b on Windows7 64bit. I have installed OpenMEEG.exe but I had this warning: *Warning! PATH too long installer unable to modify PATH!* So I tried through OpenMEEG.tar.gz, but when I try to follow the tutorial I have this warning: *'om_assemble' is not recognized as an internal or external command, * *operable program or batch file.* I added all OpenMeeg folder with subfolders to the Matlab search path, but If I write *system ('om_assemble') *the answer is the expected one only when I am inside the folder /bin. Any suggestions? Thanks in advance. Thank you, Stefano -- Stefano Caschera, PhD Student *Neuroelectrical Imaging and BCI lab* Fondazione Santa Lucia, IRCCS Via Ardeatina, 306 I-00179, Rome, Italy *Department of Computer, Control, andManagement Engineering "Antonio Ruberti"* Sapienza, University of Rome V. Ariosto, 25 00185, Rome, Italy Tel +39 06 5150 1165 Email: stefano.caschera at uniroma1.it caschera at dis.uniroma1.it s.caschera at hsantalucia.it ___________________________________________________ -------------- next part -------------- An HTML attachment was scrubbed... URL: From e.patai at ucl.ac.uk Tue Oct 24 16:57:15 2017 From: e.patai at ucl.ac.uk (Zita Eva Patai) Date: Tue, 24 Oct 2017 15:57:15 +0100 Subject: [FieldTrip] matlab nearest == FT nearest Message-ID: Hi Fters Has anyone else had any issues with using *nearest*? It is now a FT function and a Matlab 2016 one, so it often defaults to the Matlab version, which is unhelpful. If I set the path to the FT version, sometimes Matlab has a spasm. Any input would be helpful, z -- Eva Zita Patai, DPhil Postdoctoral Researcher Institute of Behavioural Neuroscience UCL -------------- next part -------------- An HTML attachment was scrubbed... URL: From Max.Cantor at colorado.edu Thu Oct 26 20:12:45 2017 From: Max.Cantor at colorado.edu (Max Cantor) Date: Thu, 26 Oct 2017 12:12:45 -0600 Subject: [FieldTrip] Comparing/contrasting ft_freqstatistics and eeglab's bootstat Message-ID: Hi, This is not a fieldtrip question per se, but I'm doing something in eeglab and I was wondering if anybody could comment on whether what I'm doing is comparable to fieldtrip's cluster permutation statistic. I'm attempting to create a statistical mask for an event-related spectral perturbation array (specifically a morlet wavelet ersp). The dimensions of the ersp are log-scaled frequency (and where number of cycles increases as frequency increases), samples, and channels. This matrix is the grand average across subjects, and the difference between two conditions. For each subject, each channel, and each condition, the ersps were baselined. In other words, the data are differences in power from baseline and between conditions, in units of decibels. I run the following inputs through bootstat: * [rsignif rbot] = bootstat(permute(g_ersp, [2,1,3]), 'mean(arg1,3);', 'alpha', 0.01, 'dimaccu', 2, 'naccu', 1000);* Where the rsignif output is a freq x 2 array which I use as the statistical mask, g_ersp is the ersp matrix I've been referring to, 'mean(arg1,3)' is the function, alpha is alpha, dimaccu is the dimension to shuffle, and naccu is the number times to reshuffle. This averages across channels (the channels are an ROI so this is what I want), shuffles across samples 1000 times, and tests for significance at alpha = 0.01. It is not testing against a baseline as I understand ft_freqstatistics to do. I use rsignif as an ersp statistical mask, and when I included the baseline vector in bootstat, it failed to mask anything. I think this is because I had baselined the ersp prior to the statistic, so literally any power tested against an empty baseline window was going to be significant. Running it in this way without testing against a baseline, I get "sensible-looking" maskings, but it would be nice to get external confirmation that what I'm doing is methodologically sound, and that I am correctly interpreting my statistic conceptually. I have used ft_freqstatistics in the past and would like to frame this bootstat statistic in a similar manner, which is why I'm asking here. Also, if I am misunderstanding my statistic, advice either on how to properly implement ft_freqstatistics-like cluster permutation statistics in this bootstat function, or alternatively how to convert my ersp matrix in such a way as to be usable with ft_freqstatistics, would be appreciated. Thanks, Max -- Max Cantor Graduate Student Cognitive Neuroscience of Language Lab University of Colorado Boulder -------------- next part -------------- An HTML attachment was scrubbed... URL: From o.jensen at bham.ac.uk Sat Oct 28 09:40:13 2017 From: o.jensen at bham.ac.uk (Ole Jensen) Date: Sat, 28 Oct 2017 08:40:13 +0100 Subject: [FieldTrip] 3 postdoc positions / U of Birmingham Message-ID: <7a17da5d-34c3-ad37-0a78-c67c68bfde5b@bham.ac.uk> Dear all, I have 3 postdoc openings funded by a Wellcome Trust Investigator Award. See here for more information. Applications can be submitted from Nov 1 to 16. ==== Applications are invited for a full-time Postdoctoral Research Fellow to work at the Centre for Human Brain Health, University of Birmingham. The successful candidate will work on a project aimed at studying the role of brain oscillations in relation to perception, attention and memory. This will be done in the group of Prof. Ole Jensen. The methodological focus of the project is MEG recordings and re-analysis of existing animal data. The candidate will be making use of the newly to be installed MEG system at the University of Birmingham and help to develop the laboratory. The candidate will be involved in all stages of designing and conducting experiments, analysing electrophysiological data. The candidate will be expected to prepare the results for high quality academic publications, to present at national and international academic conferences, and to engage in public activities. All research will be conducted in the spirit of open science. Furthermore the candidates will take part in supervision students and contribute to the Centre for Human Brain Health through leadership. === Feel free to write me for details. All the best, Ole -- Prof. Ole Jensen Centre for Human Brain Health, University of Birmingham http://www.neuosc.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From fereshte.ramezani at gmail.com Sun Oct 29 07:40:56 2017 From: fereshte.ramezani at gmail.com (Fereshte) Date: Sun, 29 Oct 2017 10:10:56 +0330 Subject: [FieldTrip] Import your own segmentation for making the head model in FiledTrip Message-ID: Dear Experts, How can one import his own segmentation data ( five labels as ; skull,scalp,GM,WM and CSF) into FiledTrip to make the finite element head model ? Thank you for your attention. Regards, Fereshte -------------- next part -------------- An HTML attachment was scrubbed... URL: From yoland.savriama at helsinki.fi Mon Oct 30 12:47:27 2017 From: yoland.savriama at helsinki.fi (Savriama, Yoland F) Date: Mon, 30 Oct 2017 11:47:27 +0000 Subject: [FieldTrip] How make a heat map on a skull using 3D landmark-based geometric morphometrics? Message-ID: Dear colleagues, I use 3D landmark-based geometric morphometrics for statistical analyses of morphological variation in vertebrate skulls. I would like to know how I could create a heat map of a skull that will show the deviations from the mean shape for a given effect around each landmark. For instance, the patterns of shape changes associated with a given Principal Component that would be mapped on a 3D skull representing the average mean shape. In this case, the visualisations of these deviations will be restricted and associated only with the local landmark coordinates as opposed to a full heat map that would cover the entire skull. In fact, the exact thing I would like to achieve had been done in Figure 4A of this publication: Maga, A. M., Navarro, N., Cunningham, M. L., & Cox, T. C. (2015). Quantitative trait loci affecting the 3D skull shape and size in mouse and prioritization of candidate genes in-silico. Frontiers in physiology, 6. https://doi.org/10.3389/fphys.2015.00092 I would greatly appreciate your feedback on this matter. Best wishes, Yoland SAVRIAMA, PhD Institute of Biotechnology P.O. Box 56 (Viikinkaari 5) FIN-00014 FINLAND -------------- next part -------------- An HTML attachment was scrubbed... URL: From mathilde.bonnefond at inserm.fr Mon Oct 30 17:07:04 2017 From: mathilde.bonnefond at inserm.fr (mathilde.bonnefond at inserm.fr) Date: Mon, 30 Oct 2017 17:07:04 +0100 Subject: [FieldTrip] postdoctoral position on the mechanistic role of brain oscillations In-Reply-To: <230d4c0ddd0a8ac1e161b1dbed847738@inserm.fr> References: <230d4c0ddd0a8ac1e161b1dbed847738@inserm.fr> Message-ID: Dear all, I'm offering a post-doc position on the role of nested oscillations in brain networks communication at the Lyon Neuroscience Research Center (CRNL) starting in February (see attached document). I would be grateful if you could share the attached job offer with interested PhD students and postdocs around you. Thank you very much. Best regards, Mathilde Bonnefond ---- Mathilde Bonnefond, PhD Lyon Neuroscience Research Center (CRNL) Inserm-CNRS-University Lyon 1 Dycog group Bron, France -------------- next part -------------- A non-text attachment was scrubbed... Name: PostDocAdvertisment.pdf Type: application/pdf Size: 232069 bytes Desc: not available URL: From heqing.psy at hotmail.com Tue Oct 31 08:45:06 2017 From: heqing.psy at hotmail.com (He Qing) Date: Tue, 31 Oct 2017 07:45:06 +0000 Subject: [FieldTrip] import Curry dataset into FieldTrip Message-ID: Dear FieldTripist: I am new to FiledTrip, could someone tell me how to read Curry dataset into FieldTrip? When I call the ft_preprocessing or ft_definetrial, it is alway showing errors. ---------------------------------------------- Error using ft_notification (line 340) unsupported header format "curry_dat" Qing He -------------- next part -------------- An HTML attachment was scrubbed... URL: