Meeting Banner
Abstract #2451

A Data-Driven Framework for Removing Physiological Noise in FMRI

Nathan Churchill1, Stephen Strother2

1Medical Biophysics, University of Toronto, Toronto, Ontario, Canada; 2University of Toronto / Rotman Research Institute, Toronto, Ontario, Canada

We have developed a multivariate, data-driven framework to estimate and control physiological noise in functional MRI. This model (1) identifies and down-weights sources of pulsatile flow (e.g. arteries, sinuses and ventricles), based on a map of high-frequency spectral power, correlated with an atlas of potential artifact regions. It then (2) regresses out physiological noise present in grey matter, using an adaptation of Canonical Correlation Analysis and the spatial weights derived from Step (1). This denoising procedure consistently reduces activation false positives, and increases both prediction accuracy and reproducibility of activation maps in subsequent analyses.

Abstract PDF Poster PDF

Keywords

accuracy activation activations adapt analyses artifacts atlas bottom church components consists control controlled correction correlated correlation developed difficult directly disco distributed dress driven exhibit false flow fraction fractional frequency functional highly hill hods house identify improves indicating issue limits making manual maps maxim maximize medical mental metrics model near noise observe often optimally optimized particularly patterns physiological plotted positives power prediction prior probable procedure process quantitative reduces regression reliability removal removing requiring resting sample sensitive separate sinus sinuses sliding spatial spectral step stimuli studies subjects task threshold trail variance ventricles vessel vessels window