Skip to main content
. Author manuscript; available in PMC: 2020 Jan 1.
Published in final edited form as: Neuroimage. 2018 Sep 1;184:293–316. doi: 10.1016/j.neuroimage.2018.08.068

Table 9:

Preprocessing strategies used to explain variance in 1.5T retrospective cohort.

VTA strategy MR volumes
used
Shown in
fig. 11
Correlation (image
metric vs. motor
improvement)
GLM (+ additional covariates)
Normalization strategy R p RSME [%] R2 Overall p Significant predictors (p < 0.05)
Binary VTA overlap
Monospectral FSL FNIRT T1 Panel A 0.38 < 0.007 15.1 0.43 < 0.005 Sex
Multispectral ANTs SyN w/o subcortical refine T1 & T2 0.46 < 0.001 15.3 0.45 < 0.005 Sex
with subcortical refine T1 & T2 0.49 < 0.001 14.8 0.48 < 0.005 Sex, VTA overlap
Multispectral SPM New Segment T1 & T2 0.51 < 10−4 14.7 0.49 < 0.001 Sex, VTA overlap
Weighted E-field VTA / STN overlap
Monospectral FSL FNIRT T1 0.46 < 0.001 15.1 0.46 < 0.005 Sex
Multispectral ANTs SyN w/o subcortical refine T1 & T2 0.47 < 0.001 15.3 0.46 < 0.005 Sex
with subcortical refine T1 & T2 Panel B 0.52 < 10−4 14.9 0.48 < 0.005 Sex, E-Field overlap
Multispectral SPM New Segment T1 & T2 Panel C 0.54 < 10−4 14.5 0.51 < 0.001 Sex, E-Field overlap
Weighted streamline counts to SMA
Weighted streamline counts to SMA seeding from E-Field (based method one row above) T1 & T2 Panel D 0.53 < 10−4 14.8 0.49 < 0.005 Sex, Streamlines to SMA