Dear all,
I did some evaluations using TBSS and I experienced, that the results were dependent of the initial skeleton threshold used. If you set this threshold too low, then the skeleton will extend away towards extremes where there is too much cross-subject variability and where the nonlinear registration has not been able to attain good alignments. But what about an upper limit? Is it accaptable if somebody set this treshold to 0.4?
Thanks,
Gabor
|