Dear SPM community,
I am trying to run segmentation on one of my subjects and received the
"Warning: Matrix is close to singular or badly scaled" message. The
resulting images are not skull stripped and the tissue segmentation itself
also looks very different, almost like the image is biased incorrectly. I
have tried the recommended steps to resolve this issue; updating SPM5,
changing the origin of the structural to the AC, and checking the scale
factor in mricro. The warning/error message no longer appears, but
unfortunately the segmentation images still appear incorrect.
Does anybody have any other ideas as to how to resolve this issue?
Thanks,
Carlos Faraco
University of Georgia
|