Hi, We have been using the latest eddy patch to correct our multiband diffusion data. For all the 3 subjects that we processed today, the eddy-corrected-dwi has almost half of the top of the brain cut off. But eddy-corrected-dwi-outlier-free images looks correct. Now the question is, can we simply use the outlier -free data to do further processing? If no, what else can be done. I am running eddy on Ubuntu 16 using the centos6 patch. In the past we haven't encountered this problem and any insights will be greatly appreciated. Thank you Regards VM