 |
 |
Hi, Have you run the same dataset through the new and old versions? If it was the one you were converting from hdr/img/mat format then there might be issues in the conversion which are causing the problem. Also, what happens when you run the new melodic on the FEEDS data? All the best, Mark On 5 Sep 2008, at 02:18, Xuelin Cui wrote: > hi folks: > > I run new MELODIC in FSL (4.1.0) to my 4D NIFTI fMRI data. But the > output files seem change a lot > > 1. The running time seems much longer. It costs me around 2 hours to > do ICA analysis for one run while it was only 10-15 minutes before. > > 2. The number of independent components decreases some. I used to > get around 20-30 independent components, but now I only get 11. > > 3. The time course looks different. In old version of MELODIC, I can > easily see up and down, which corresponds to activation or resting, > in time course. But in new MELODIC results, I only see some spikes, > which doesn't make much sense to me. > > So I am not sure if I did wrong somewhere, any comments on that? > > Thanks > > Xuelin > > > > **************************************** > Xuelin Cui > Department of Electrical Engineering > University of Hawaii-Manoa > Honolulu HI 96822 > > Tel: 1-808-956-0856 > Email: [log in to unmask] > ****************************************