As a newbie to SPM and VBM, I have been following John's detailed tutorial on how to perform VBM with SPM. 

In the segmentation step, I follow John's advice, leaving many settings as default, and it works perfectly fine for most of the time. It fails only once for all my ~1000 subjects.

I have attached the screenshots of the segmented GM, WM, and CSV for that failure case. Could anyone kindly help analyze the reason why the results look like that? It would be wonderful if one could also point me which parameter I should fine tune in order to fix this.

Thank you all very much in advance!

Marris