Dear Haobo,

Thank you for your interest in PANDA.

When you display results, PANDA will create a *_SigIndex.nii.gz, which is a mask of significant regions. In this map, if there are four regions, the value of each region will be 1, 2, 3, 4.
You should first binarize this mask image, to make is a binary map.
You can use this binary map to multiply the t map to extract the t values of significant voxels.
Then you can use other softwares to find the peak voxels, for example, REST, xjview.

If you have any problems, you can contact me. You can send me email, so that I can receive the message immediately.

Please cite PANDA's paper when you writes paper, thank you very much.

Best wishes
-
Zaixu