A Deep Pattern Recognition Approach for Inferring Respiratory Volume Fluctuations from fMRI Data

Abstract

Functional magnetic resonance imaging (fMRI) is one of the most widely used non-invasive techniques for investigating human brain activity. Yet, in addition to local neural activity, fMRI signals can be substantially influenced by non-local physiological effects stemming from processes such as slow changes in respiratory volume (RV) over time. While external monitoring of respiration is currently relied upon for quantifying RV and reducing its effects during fMRI scans, these measurements are not always available or of sufficient quality. Here, we propose an end-to-end procedure for modeling fMRI effects linked with RV, in the common scenario of missing respiration data. We compare the performance of multiple deep learning models in reconstructing missing RV data based on fMRI spatiotemporal patterns. Finally, we demonstrate how the inference of missing RV data may improve the quality of resting-state fMRI analysis by directly accounting for signal variations associated with slow changes in the depth of breathing over time.