Single subject fMRI test-retest reliability metrics and confounding factors

► We have scanned 10 volunteers twice using four different fMRI paradigms. ► We have measured reliability of timeseries, unthresholded and thresholded t maps. ► Additionally we measured scanner noise, subjects motion and coregistration errors. ► Task and subject motion had major contribution to reliability. ► Scanner noise and coregistration errors had little contribution to reliability.