A lag structure in the fMRI signal, created by a combination of the low-frequency oscillation and perfusion delay, has been suggested as being a noise component, but its removal has not been established as a denoising approach. Using an HCP task-fMRI dataset, we report that the removal of this lag structure, or deperfusioning, has a unique effect on fMRI group analysis both qualitatively and quantitatively. Increased sensitivity of the treatment in some condition was emphasized by switching from volume- to surface-based analysis, suggesting concomitant improvement of specificity. Unfavorable effect was also suggested depending on the task condition, warranting further investigation.
This abstract and the presentation materials are available to members only; a login is required.