External Email - Use Caution
I have created a volume which is supposed to have the cortical surface orientation with respect to b0 at each voxel. To do this I got the 6 frame overlay file from mris_convert. Then I isolated the first 3 values from each row in Matlab to create an Nx3 matrix, then from this I calculated the angle between each row vector and the b0 vector (which was just [0 1 0]) using the usual angle between two vectors formula. The result was an Nx1 matrix.I then saved it as an .mgh surface file (using save_mgh) and projected it back onto the original volume using surf2vol. I believe that I did this correctly, however the professor that I am developing this method for told me he wants me to verify that the results are correct. He was not very specific in how to do this, all he told me was that I could use volumes of the individual scanner x,y,z angles to help me figure out if my b0 results are right. What I assume he meant is: since the b0 axis goes vertically down the middle of the brain when in coronal view I can tell if the b0 angles are correct by seeing where the cortical surface is with respect to the z scanner direction and seeing if the b0 angle at the corresponding location makes sense. My problem is that I can’t tell if my results make sense. In theory I believe that at the locations where the cortex appears to be orthogonal to the B0 (z scanner) axis, the overall B0 angle of the cortical surface should be closer to 90 degrees. I still don’t understand what pattern I should be seeing in the values of the z scanner volume compared to the b0 angles volume. Comparing my z scanner volume results to the B0 angle volume the only pattern I can see is that the cortical surface with respect to B0 is closer to 90 degrees when the z scanner volume is closer to 0 degrees. No B0 angle I got is actually greater than 89 though. I can’t tell if this is the right pattern or enough to verify the results in the way the professor I’m working for wants. Any help here would be GREATLY appreciated. So far I have been able to get pretty far in teaching myself Freesurfer, but the some of the underlying mathematics about how to verify my results are unclear. I can send pictures of my results if someone wants to look at them.