HI Octavian
There are a set of N priors that are pruned to give a smaller subset.
In the standard approach these N priors start off at predefined cortical locations.
You can however have random location for these N priors (or vary N).
see section
38.3.5 Other MSP options
from the SPM12 manual.
best
gareth
________________________________________
From: SPM (Statistical Parametric Mapping) <[log in to unmask]> on behalf of Octavian Lie <[log in to unmask]>
Sent: 04 January 2014 07:03
To: [log in to unmask]
Subject: [SPM] Inverse solution simulations
Dear All,
A rookie question:
Are the spatial coordinates of multiple sparse priors (n= no of source
priors/hemisphere) fixed or resulting from best model selection
(variable coordinates, constant n)?
If the latter is true, does the latter minimize the inverse crime
regarding the source space selection when using synthetic data for
inverse solution simulations, or one still has to select different
discretizations (different n) of the source space?
Thank you,
Octavian
|