For doing surface-based statistical analysis, I imagine you would want a program like AlphaSim that gives the probability of false positives in terms of node clusters (or surface area) rather than voxel clusters. By any chance, is such a program available or in the works?
thanks,
Rutvik