No subject

[Robert Oostenveld]

Hi Sameer,

One small comment to start with: please use a subject in your emails
to the list. That makes it easier to keep track of discussions and
threads.

On 28 Feb 2007, at 21:59, Sameer Walawalkar wrote:
> I am currently experimenting with beamformers and was playing with
> http://www2.ru.nl/fcdonders/fieldtrip/doku.php?
> id=fieldtrip:documentation:use_your_own_forward_leadfield_model_in_an_
> inverse_beamformer_computation
> to generate the grid.

You have 306ch Neuromag data, and I don't recall exactly whether that
has been fully tested for forward computations. Some background: at
each of the 102 locations of each "chip" you have 3 channels: two of
those have 2 coils, and one is a magnetometer with only one coil.

Please have a look at http://www2.ru.nl/fcdonders/fieldtrip/doku.php?
id=fieldtrip:documentation:frequently_asked_questions#how_are_electrodes
_magnetometers_or_gradiometers_described
for the description of "grad". I would expect in your case that
  grad.pnt and ori   102*2+102*2+102= 510 rows long, each with 3 columns
  grad.tra           306 x 510

The gradiometer coils are shifted w.r.t. the center of the "chip" by
a small amount. I think that currently the baseline (i.e. the
distance between the coils) is hardcoded. Joachim Gross once gave me
code for that, it is included in feldtrip/private/fif2grad. Please
look at it and see whether it applies to your 306ch gradiometer
design as well.

>
> I did the following
>
>
> vol   = [];
> vol.r = 12 * [0.88 0.92 1.00]; % radii of spheres, the head radius
> is 12 cm
> vol.c = [1 1/80 1];            % conductivity
> vol.o = [0 0 0];              % center of sphere

this indicates a 4-shell concentric-sphere EEG model. It does not
make sense for MEG, since in MEG teh conductivity of the different
layers in the spherical model does not affect the field distribution.

> % compute the leadfields that can be used for the beamforming
> cfg            = [];
> cfg.vol        = vol;
> cfg.resolution = 2;  % same unit as above, i.e. in cm
> grid = prepare_leadfield(cfg, dataPre);
>
> (I expect prepare_leadfield to pick up grad definitions from the data)
>
> ??? Error using ==> fieldtrip/private/compute_leadfield
> different number of spheres for the radius and origin

FT works with MEG either with a single sphere or multisphere models,
i.e.
   vol.r = 12
   vol.o = [0 0 5]
This origin is a better guess for the center of the sphere, remember
that in the MEG case you do not move the sensors towards the head,
but the position of the sensors (coils) is expressed with respect to
the head coordinate system. In case of EEG, it is more common to use
standard electrodes that are moved towards the head. You can use
HEADMODELPLOT to look at the geometry of the sensors and head model.

A multisphere model contains a sinsgle sphere for each sensor.
Actually, for the forward computation it should contain a single
sphere for each coil. Multisphere models currently are not
implemented for Neuromag data.

The reason for the error is that FT is confused about it being not a
single-sphere, but also not a multi-sphere model. Also your later
errors are explained by that. Please use a single sphere model. Or
alternatively you can use a realistic single-shell (Nolte) model,
which is more accurate than the multi-sphere models, but I suggest
first to try around with a single sphere. The most important thing
for you to check is whether the grad structure is geometrically
correct. Lauri Parkonen will probably be able to help you with
checking that.

best regards,
Robert


> Error in ==> prepare_leadfield at 209
>   grid.leadfield{dipindx} = compute_leadfield(grid.pos(dipindx,:),
> sens, vol, 'reducerank', cfg.reducerank, 'normalize', cfg.normalize);
>
> Error in ==> SomeMaster at 15
> grid = prepare_leadfield(cfg, dataPre);
>
>
>
> To get around this error, I change vol to
>
> vol.o = [0 0 0 ; 0 0 0 ; 0 0 0  ];
>
> ??? Error using ==> fieldtrip/private/compute_leadfield
> number of spheres is not equal to the number of coils
>
> Error in ==> prepare_leadfield at 209
>   grid.leadfield{dipindx} = compute_leadfield(grid.pos(dipindx,:),
> sens, vol, 'reducerank', cfg.reducerank, 'normalize', cfg.normalize);
>
> Error in ==> SomeMaster at 15
> grid = prepare_leadfield(cfg, dataPre);
>
>
> I thought this might be due to the 4 HPI coils, so I changed vol to
> vol.r = 12 * [0.88 0.92 0.96 1.00];
> vol.o = [0 0 0 ; 0 0 0 ; 0 0 0 ; 0 0 0  ];
>
> but it was of no use.
>
> Any suggestions?
>
> Thanks,
> sameer
>