[FieldTrip] Methods discussion: separating neural coupling from EMG in paired depth and external EEG

[נתי שטרן]

Dear FieldTrip community,

I would welcome methodological feedback on a proposed study recording
depth electrophysiology and external EEG simultaneously in awake
animals after recovery from implantation, with synchronized video and
EMG.

The question is whether a signal preceding aggressive action can be
recovered in external EEG, with a warning interval that ends before
the first annotated movement. The proposed framework covers control,
ASD-relevant, SCZ-relevant, and combined experimental conditions,
whose construct validity would need separate assessment.

For a FieldTrip-based analysis, which controls would you prioritize to
distinguish genuine depth-to-external coupling from a shared
reference, volume conduction, and muscle activity? The draft considers
reference sensitivity, imaginary coherence or weighted phase-lag
measures where appropriate, matched non-attack movement, causal
preprocessing, and held-out animals. It treats cross-modal association
and prospective forecasting as separate validation steps.

The manuscript is a hypothesis and proposed validation framework, with
no new animal data, trained predictor, or validated aggression-warning
signal. It does not infer dangerousness from an ASD or schizophrenia
diagnosis. ChatGPT assisted with literature discovery and drafting, as
disclosed in the manuscript.

Manuscript: https://github.com/netanelcyber/psi-qeeg-bis/blob/docs/pre-attack-eeg-validation-20261025/docs/research/PRE_ATTACK_EEG_VALIDATION_2026.md

Best regards,
Netanel Stern
Independent researcher, Israel
nsh531 at gmail.com