Meeting Banner
Abstract #2293

Reliability of fMRI: Are Group Means Really Representative?

Tynan Stevens1, Steven Beyea2, Ryan D'Arcy3, David Clarke1

1Dalhousie University, Halifax, Nova Scotia, Canada; 2NRC, Halifax, Nova Scotia, Canada; 3Frasier Health Authority, Surrey, British Columbia, Canada

The Rombouts overlap coefficient (Roverlap) is used extensively in fMRI reliability research. While it is widely recognized that this coefficient depends on analysis threshold, existing studies do not agree on the precise relationship between Roverlap and threshold, and so no clear strategy has been employed across studies. We argue that individual variability in the Roverlap-threshold relationship reduces the information available in group level Roverlap results. This variability likely contributes to the inconsistent findings in the literature to date. We thus suggest that control of reliability is best conducted at the individual-level, as group trends may not be truly representative.

Keywords

absence activated activation active activity affine allowing alternating although anatomical apart authors averaging block blocks boxcar brain classified clear coefficient coefficients collected combination common complex consistent control convolved corrected count dataset datasets days decrease decreased decreases decreasing default dependence describing distinct distributed effectively eight emerged especially example explain finger fixed function functional furthermore greatest health healthy help highly identify illustration increasing individual individuals intensity interspersed investigates known likely linear literature local location lost majority males maps matrix maximal measure measuring monotonically motion normalization nova occurs often optimal others overall overlap pace pair pairs paper participant patterns performance perpendicular practice previous produce programmed projector python rapidly really reduce registered regressors relationship reliability reliably report reported representative reproducibility respectively response rest retest scanner scanning separate session sessions significance slices slowest smoothed space statistical stimuli strategy strength structural subjects surrey tapping task terms threshold thresholds timing tools typically unique variability variable volume volumes volunteers