Improving power in functional magnetic resonance imaging by moving beyond cluster-level inference

Abstract

Inference in neuroimaging typically occurs at the level of focal brain areas or circuits. Yet, increasingly, well-powered studies paint a much richer picture of broad-scale effects distributed throughout the brain, suggesting that many focal reports may only reflect the tip of the iceberg of underlying effects. How focal versus broad-scale perspectives influence the inferences we make has not yet been comprehensively evaluated using real data. Here, we compare sensitivity and specificity across procedures representing multiple levels of inference using an empirical benchmarking procedure that resamples task-based connectomes from the Human Connectome Project dataset (∼1,000 subjects, 7 tasks, 3 resampling group sizes, 7 inferential procedures). Only broad-scale (network and whole brain) procedures obtained the traditional 80% statistical power level to detect an average effect, reflecting >20% more statistical power than focal (edge and cluster) procedures. Power also increased substantially for false discovery rate– compared with familywise error rate–controlling procedures. The downsides are fairly limited; the loss in specificity for broad-scale and FDR procedures was relatively modest compared to the gains in power. Furthermore, the broad-scale methods we introduce are simple, fast, and easy to use, providing a straightforward starting point for researchers. This also points to the promise of more sophisticated broad-scale methods for not only functional connectivity but also related fields, including task-based activation. Altogether, this work demonstrates that shifting the scale of inference and choosing FDR control are both immediately attainable and can help remedy the issues with statistical power plaguing typical studies in the field.

Co-fluctuations of neural activity define intra-V1 networks related to perceptual organization

“…We confirmed these results obtained with edge-based analysis by using a constrained network-based statistic (cNBS), which has increased power compared with other methods for testing hypotheses about networks (Noble et al, 2022). This approach requires independently defined subnetworks to avoid circularity, which is achieved with the connSubsp groups.…”

“…FDRs were estimated for both tails of the t-value distribution by using mixture modelling (Bielczyk et al, 2018; Gorgolewski et al, 2012). Edgewise analysis has the advantage of greater spatial precision but also less statistical power (Noble et al, 2022). Measures related to topology were derived from the networks sparsified with the FDR thresholds.…”

“…Furthermore, we propose a novel approach to performing intra-V1 connectivity analysis (pipeline in Figure 2). The goal was to identify reliable differences in intra-V1 correlation matrices across conditions, employing edge-based and network-based tests (Chung et al, 2021; Noble et al, 2022). Taking advantage of the close correspondence between V1 cortical anatomy and retinotopy (Benson et al, 2012; Benson & Winawer, 2018), we mapped onto the visual field those links between cortical sites that differed consistently between stimulus classes.…”

Topological Cluster Statistic (TCS): Towards structural-connectivity-guided fMRI cluster enhancement

“…In addition to the improvements in detecting localized activation patterns, TCS provides additional information by linking spatially disjoint effects into a unified anatomical network. This is a crucial feature that addresses potential limitations of existing localized inference approaches 22 . As TCS uses a high-resolution anatomical topology, it can provide rich anatomical information about the underlying white-matter fibers connecting an activation map.…”

“…Anatomical connections support communication between distant brain areas, thus enabling the coordinated activity of non-contiguous groups of gray matter voxels 20,21 . While some have explored the impact of grouping non-contiguous areas, there have not yet been any attempts to group areas based on the known underlying anatomy 22 .…”