Given that you have 200 brains to work with... you might consider a semi-automated procedure as briefly outlined below:

1. Obtain the MNI template or other template for the CC. (or u could create your own by creating a probabilistic map of the CC for your 200 brains in talairach space).
2. Align this CC mask into talairach space.
3. Dilate this talairached mask to make it a bit more inclusive.
4. For each subject, transform this talairached mask into the subject's original space.
5. Apply mask onto anatomical data and count voxels that lie within mask and exceed some intensity threshold (which you could determine from finding the white matter peak in the histogram of the anatomical data, be sure to run 3dUniformize before creating the histogram).
6. Compute volume based on the voxel count.

As I type this I realize that this might take as long to implement and test as manually drawing the ROIs, but on the other hand it would be a 'tad' less painful. :)


good luck,
-shruti