Examining reliability of swallowing activation with Partial Least Squares and SimulScan

Motivation: SimulScan enables the joint imaging of dynamic swallowing along with the central control from functional MRI. Partial Least Squares (PLS) analysis applied to SimulScan data extracts latent variables that describe fundamental swallowing brain activation. Goal(s): We evaluate the reliability of PLS analysis applied to SimulScan data to identify correlated regions of brain activity and dynamic swallowing motions. Approach: We use our memory-efficient implementation of PLS to extract latent variable masks from dynamic and functional imaging of swallowing tasks in a repeated scan. Results: The PLS method extracts dynamic and functional imaging latent variable masks that reliably capture fundamental swallowing function. Impact: Partial least squares analysis of SimulScan data can reliably identify latent variables that capture the fundamental brain function of swallowing. This approach will enable the in-depth study of the healthy and disordered swallowing mechanism in age and disease.