Hello,
I am working on an experiment where subjects perform a pointing task. For each subject, there are 6 different conditions, each corresponding to a direction, with several trials for each direction.
When estimating sources in Brainstorm (2018 version), is it better to compute sources using only one condition (one direction) per subject, or should I use all six conditions (all directions) from each subject to estimate the sources?
What is best to do?
Thank you for your advice!
The source estimation depends on the Head model and the Noise covariance.
The head model will be the same for all conditions if there was not change in the electrode positions across conditions.
The noise covariance will be the same for all conditions if it is computed from the same baseline period all the conditions
Thus if the head model and the noise covariance are the same for all your conditions, you only need to compute sources once, this will create a Inversion kernel (
).
Check the these tutorials (in this order) on head-model, noise covariance and source estimation
In those the tutorials the inversion kernel is compute only once per session