For first level analysis I'd recommend using the output from standard  
melodic. The reason why in a tica analysis you have less components  
simply is that effects which might be relevant at the within-subject  
level (due to them being large relative to the within subject  
variance) are not necessarily significant at the mixed-effects level.  
Nevertheless, in order to get best de-noising at the subject level you  
should be using the relevant subject level ICA (and not the group ICA).

On 17 Aug 2009, at 19:03, Matthew Ward wrote:

> Hello everyone.  I have a question about using the output from a
> multi-session Tensor_ICA analysis.  I understand how to filter  
> individual 4D
> datasets using single session Melodic data exploration and the  
> resultant
> components and how to reanalyze the filtered datasets in a lower level
> analysis, but is there a way to use multi-session tensor_ICA  
> 'melodic_mix'
> files to filter the 4D datasets used in a higher level FEAT  
> analysis?  And,
> if so, is the 'melodic_mix' file or the 'melodic_FTmix' file found  
> in the
> *.gica directory the proper one to use?  Or is it a matter of just  
> using the
> individually filtered data in the higher level FEAT analysis?  I ask  
> because
> the number of components in the multi session output is considerably  
> less
> than the components associated with the individual MELODIC outputs.   
> Thank
> you for your time and thought on the matter.
> Best,
> Matt

Christian F. Beckmann, DPhil
Senior Lecturer, Clinical Neuroscience Department
Division of Neuroscience and Mental Health
Imperial College London, Hammersmith Campus
Rm 419, Burlington Danes Bldg, Du Cane Road, London W12 0NN, UK
Tel.: +44 (0)20 7594 6685   ---   Fax: +44 (0)20 7594 6548
Email: [log in to unmask]

Senior Research Fellow, FMRIB Centre
University of Oxford
JR Hospital - Oxford OX3 9DU
Tel.: +44 (0)1865 222551 --- Fax: +44 (0)1865 222717
Email: [log in to unmask]