sburns.org Sometimes auto correct just isn't helpful
Sent from my iPhone
> On Jun 11, 2016, at 5:16 PM, Robert Kraft <[log in to unmask]> wrote:
>
> While Fsl has great tutorials. I also found the site suburbs.org to be very helpful. Step by step instructions provided using Fsl tools.
>
> Sent from my iPhone
>
>> On Jun 11, 2016, at 4:37 PM, Lance Stevens <[log in to unmask]> wrote:
>>
>> Re the tractography practical, I am uncertain as to how to calculate the probability of connection between seed mask (here the right thalamus) and target cortical mask. I don’t have much experience with FSL, so I apologize in advance for the likely simplistic nature of this question.
>>
>> Specifically, in this segment (Connectivity-based segmentation: http://fsl.fmrib.ox.ac.uk/fslcourse/lectures/practicals/fdt2/index.html#conseg ), we are asked to “compare the spatial distributions and probabilities of connection from different thalamic voxels to different cortical target zones” for this subject.
>>
>> Looking at the output (seeds_to_* files) with fslstats (e.g. fslstats seeds_to_MASK_average_M1_right.nii.gz -V 716 5728.000000; fslstats seeds_to_MASK_average_temporal_right.nii.gz -V 1109 8872.0000000) the voxel values showing the highest number of samples reaching the target mask come from the seeds_to_MASK_average_temporal_right.nii.gz image; this image is listed as the 7th out of 7 after running the find_the_biggest call. I would have thought that (reciprocal) connections between the right thalamus and M1 would be greater?
>>
>> It’s apparent that the pro_thresh command is used to arrive at the probabilities of connection between the seed mask (proj_thresh <lots of volumes/surfaces> threshold), but what should the threshold be 50, 100, 500?. I assume this will be (fractions) out of 5000? If example usage for this data (fdt2/subj1.bedpostX/THAL2CTX_right) could be provided it would be greatly appreciated.
>>
>> Best,
>> Lance
|