Sleep Scoring Automation Via Large Scale Machine Learning

Chris R. Fernandez, Samuel J. Rusk, Nick J. Glattard, and Mehdi Shokoueinejad

In this work, we present a large-scale machine learning analysis of the multi-site, 5793 patient SHHS dataset. We argue for the benefits of a rigorous scoring based framework for estimating OSA diagnostic parameters, and demonstrate state of the art performance in SDB event classification on SHHS polysomnography data, driven largely by recent advancements in high performance computing and the rapid proliferation of evolving machine learning techniques.