Hi Stephen,

the TFE is probably fine, or a good quality mp-rage. My only concern with the 8 channel is how uniform the B1 field is. Do you see large drop-offs near the top of the brain? We currently don't support multi-spectral inputs for most of what we do, although we do have some prototypes that do.

cheers, Bruce

On Thu, 16 Jul 2009, Stephen Towler wrote:

Hi freesurfers,

Are many groups getting good results from single 1mm isovoxel T1 inputs? Anyone experimented with multispectral input?

We're acquiring a number of participants on a Philips Achieva 2.5 system with 8-ch SENSE coil. Good scans, very little movement. I realize that the selection of freesurfer inputs and averaging is empirically driven and idiosyncratic to each scanner/sequence/coil combination, but I just want to solicit any thoughts about possible benefit from non-standard sequences. Due to other priorities we are unable to collect two identical T1s in each study, but this is what we do get that might aid automated structural segmentation:

• 3D T1 TFE (1 mm3, good contrast among gray, white, and CSF)
• 3D T2 FLAIR (1 mm3, high gm/csf contrast but no gm/wm contrast, except for

wm hyperintensities)

• 3D T1 "FGATIR" (.75 mm3, very high gm/wm contrast but very low gm/csf

contrast)

This last sequence is the one I'm most curious about: might the very high gm/wm contrast be used to help freesurfer's segmentation? Image attached, showing a participant's standard 3D T1 TFE on the right and the high g/w contrast T1 on the left. This paper describes the FGATIR sequence:

Sudhyadhom, A., Haq, I. U., Foote, K. D., Okun, M. S., & Bova, F. J. (2009). A high resolution and high contrast MRI for differentiation of subcortical structures for DBS targeting: The Fast Gray Matter Acquisition T1 Inversion Recovery (FGATIR). NeuroImage. doi: 10.1016/j.neuroimage.2009.04.018.

Thanks! Stephen