 |
 |
Hi Eugene, I am working on processing "Quantitative Susceptibility Mapping" images. The pipeline that I used output COPE images. I want to run whole brain voxel wise analysis using randomise. When I merged the COPE images and fed them into Randomise, this output weird statistical maps. I could not understand the results. I suspect! that the positive and negative values in these COPE images were not handled correctly in randomise, so I am hoping to convert these images to images has positive values. Hi John - You'll need to explain for what reason you want nly positive values - what type of images are you looking for / what do you want to do with them? Cheers Eugene On 11 April 2018 at 18:35, John anderson <[log in to unmask]> wrote: Dear FSL experts, contrast of parameter estimate (COPE) represent the final images in FEAT fMRI data analysis, I understand that are calculated by performing simple algebra on the EV parameter estimates. For example, for the contrasts 1 -1, the contrast parameter estimate is calculated by subtracting the parameter estimate for EV 2 from the parameter estimate for EV1. The COPE image will show regions in the brain in which this difference is large. (http://fsl.fmrib.ox.ac.uk/fslcourse/lectures/practicals/feat2/index.html). My questions is: The cope images contain negative and positive values. Is there any way to convert these images to another type of images that contain only positive values? Thank you for any clarification john.