The impact of T1 versus EPI spatial normalization templates for fMRI data analyses

Abstract

Spatial normalization of brains to a standardized space is a widely used approach for group studies in functional magnetic resonance imaging (fMRI) data. Commonly used template-based approaches are complicated by signal dropout and distortions in echo planar imaging (EPI) data. The most widely used software packages implement two common template-based strategies: (1) affine transformation of the EPI data to an EPI template followed by nonlinear registration to an EPI template (EPInorm) and (2) affine transformation of the EPI data to the anatomic image for a given subject, followed by nonlinear registration of the anatomic data to an anatomic template, which produces a transformation that is applied to the EPI data (T1norm). EPI distortion correction can be used to adjust for geometric distortion of EPI relative to the T1 images. However, in practice, this EPI distortion correction step is often skipped. We compare these template-based strategies empirically in four large datasets. We find that the EPInorm approach consistently shows reduced variability across subjects, especially in the case when distortion correction is not applied. EPInorm also shows lower estimates for coregistration distances among subjects (i.e., within-dataset similarity is higher). Finally, the EPInorm approach shows higher T values in a task-based dataset. Thus, the EPInorm approach appears to amplify the power of the sample compared to the T1norm approach when not using distortion correction (i.e., the EPInorm boosts the effective sample size by 12-25%). In sum, these results argue for the use of EPInorm over the T1norm when no distortion correction is used. Hum Brain Mapp 38:5331-5342, 2017. © 2017 The Authors Human Brain Mapping Published by Wiley Periodicals, Inc.

Keywords: coregistration; echo planar image; fMRI; spatial normalization.

Figure 1

Side‐by‐side comparison of EPInorm and T1norm approaches for a single subject transparently overlaid on the T1 image from that subject. The T1norm process is unable to compensate for distortions throughout the brain, which are not present in the T1 scan (blue circles). [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 2

Schematic of the EPInorm and T1norm approaches. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 3

T1norm versus EPInorm in Experiment 1a: (a, left) T1 norm voxelwise subject standard deviation, (a, middle) EPInorm voxelwise subject standard deviation, (a, right) difference (T1norm–EPInorm). (b) Violin plot of the voxels showing a subject standard deviation of over 25, T1norm shows a higher whole brain mean. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 4

T1norm versus EPInorm in Experiment 1b: (a, left) T1 norm voxelwise subject standard deviation, (a, middle) EPInorm voxelwise subject standard deviation, (a, right) difference (T1norm–EPInorm). (b) Violin plot of the voxels showing a subject standard deviation of over 25, T1norm shows a higher whole brain mean. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 5

T1norm versus EPInorm in UC Boulder dataset: (left) T1 norm voxelwise subject standard deviation, (middle) EPInorm voxelwise subject standard deviation, (right) difference (T1norm–EPInorm). [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 6

Alignment of first image for each participant relative to a random KKI subject. The EPInorm approach showed significantly more similarity (P < 0.05) among subjects in alignment relative to the T1norm approach (Wilcoxon signed‐rank test, V = 631, P = $1.59\times {10}^{-13}$). [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 7

T1norm versus EPInorm in ABIDE dataset: (a, left) T1 norm voxelwise subject standard deviation, (a, middle) EPInorm voxelwise subject standard deviation, (a, right) difference (T1norm–EPInorm). (b) Violin plot of the voxels showing a subject standard deviation of over 25, T1norm is clearly much higher. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 8

T values corresponding to false alarms versus hits for the go/no‐go task without distortion correction for (a) T1norm and (b) EPInorm. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 9

T values corresponding to false alarms versus hits for the go/no‐go task with distortion correction for (a) T1norm and (b) EPInorm. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 10

Single‐subject T values for the four cases (T1norm, EPInorm, T1norm with distortion correction, and EPInorm with distortion correction): EPInorm shows the highest mean T‐values and T1norm without distortion correction is significantly lower than the other three approaches. [Color figure can be viewed at http://wileyonlinelibrary.com]