We have three 7T scans of the same person done on separate days. After running all three T1 images using the "recon-all" function, we obtain thalamic volumes. The volumes varied by up to 25% which seems quite drastic for the same person. Is there a way to decrease this variability? Are there some parameters to change for 7T? Would the subcortical segmentation work better with a T2 or DTI image?