We currently acquire two sagittal T1s (136 slices) in our protocol so that we can take advantage of freesurfer's motion correction algorithm; however, we would like to add a coronal T1 (196 slices). In order to cut down on scan time, we are also considering dropping one of the the sagittal T1s and using one sagittal and one coronal T1 as inputs for freesurfer recon-all.

When I specified these sagittal and coronal inputs for the recon-all for our test subject, we received the "ERROR: inputs have mismatched dimensions!" My main questions are the following:

If we do cut one of the sagittal T1 scans, 

1) Would it be possible to use one sagittal and one coronal T1 as inputs? The sequences are identical except for the acquisition plane and number of slices.

2) If it is possible (perhaps by registering with mri_robust_register), what would be the correct procedure for matching the dimensions of these two scans?

3) Are there any other considerations or possible problems with this and does this approach have merit (i.e. would it be advisable to just use one T1 input instead of two T1 inputs that have different acquisition planes)?

Any insight would be greatly appreciated!