Dear Jesper In my old experiment, we acquired 13 directions DTI dataset with only half sphere acquisition. Could I use eddy to correct these old dataset ? Best Paul 2015-04-23 21:14 GMT+08:00 Jesper Andersson <[log in to unmask]> : > Dear Thais, > > > > > I did the FSL course last year (Oxford) and learned that to process the > DTI images wee need to use 'topup' and 'eddy'. However, the dataset that I > will analyse was collected quite while ago and as far as I know, it was not > acquired with 2 different phase encoding directions, so I guess I cannot > use 'topup' and 'eddy'. So, could you please help me with my queries. > > > > 1- In this case, instead of 'eddy' I will have to use 'eddy_correct' > from the FDT. >>> Is this the way forward? Any alternatives? > > you need two PE directions to run topup. It is not necessary in order to > run eddy. When you run eddy without having run topup first it means that > you only correct for eddy current distortions and subject movement (i.e. > not for susceptibility). But that is still likely to be better than to use > eddy_correct (that doesn’t attempt to correct for susceptibility either). > > Can you please give me some details about your diffusion data set, such as > number of diffusion directions, if directions were sampled on the whole or > half sphere? > > Jesper > > > > > > 2- To 'eddy_correct', I do not need to use the 'fslroi' command to > select a volume without diffusion weighting, as in 'eddy_correct' there is > the 'reference no' which already accomplish for it. >>> Am I correct? > > > > 3- For the 'bet' command, which will create a file to use on 'dtifit', > do we have to use 'fslroi' on the original DTA.nii file (to select a volume > without diffusion weighting), or can we use 'bet' directly on the original > DTA.nii file? > > > > Many thanks for your help! > > > > Regards, > > > > Thais Minett >