 |
 |
Hi Marco, Please, see below: > 1) When I use -t to threshold and build clusters, the value of -t refers > to the value of t-maps at a voxel level, right? > The -t refers to the *t*hreshold. If the input image is a z-map, then this value should be a z-score. If it's a t-map, then it's a t-statistic. If it's a p-values map, then -t is a p-value. > 2) How does FLS calculate the p-values of each cluster based on the > smoothness (--dlh) and number of voxels in the mask (--volumes)? > It uses the RFT for clusters. The probability of finding a cluster larger than a certain size is approximated by an exponential curve, that depends on the threshold that defines the clusters, on the smoothness, dimensionality, etc. It can be calculated once all the parameters are available. 3) Is this p-value of the clusters related to extension or peak or any > other features of the clusters? > It's the p-value for the size of the cluster (i.e., the cluster extent). All the best, Anderson