This is what I use for this question:
http://research.ics.aalto.fi/ica/icasso/, usually in the context of a
FASTICA runs in matlab (which is the core algorithm inside melodic).
Peace,
Matt.
On 9/26/16, 3:40 PM, "FSL - FMRIB's Software Library on behalf of George
chahine" <[log in to unmask] on behalf of [log in to unmask]> wrote:
>Hi all,
>
>One of my reviewers requested information about whether I assessed for
>stability of components in my ICA analysis. As far as I know, Melodica
>doesn't assess automatically for ICA stability. I'm just wondering if
>there is any method to assess for stability in FSL or if there is any
>study about test-retest reliability of melodica. FTR, I performed the ICA
>analysis on a total of 303 subjects, which is probably more than most
>studies out there.
>
>Thank you,
>George
|