Hi Bruce, I have done the recon-all (fs5.3) with the data collected from a 7T scanner using MP2Rage. The recon went well without problem. But the the temporal reconstruction looks clearly not correct (see attached). I tried to push the pial surface outside by editing the white matter evidenced in the attached figures, but has no effect to genera correct pial surface. I do realize that the contrast is low in the anterior temporal lobes. However, I am wondering if there are expert parameters I can use to push the pial surface into the right place. If that isn't a option, can I use some other images collected during the MP2rage scan to help the segmentation in the anterior temporal lobe? Those extract scans are: inv1_nd, inv1_phs_nd, , t1, inv2_nd, invs_phas_nd, thanks very much. Xiaomin