The study represents a deliberate, multi-disciplinary effort to use novel and timely machine learning techniques to study the relationship of health outcomes and phenotypes with the incidental information captured in cardiovascular diagnostic testing with electrocardiograms (ECGs). Medical imaging and waveform data is routinely and frequently used for diagnosis and prognostication in routine clinical care, however there is often subjectivity in interpretation, heterogeneity in application, and variance with image acquisition and quality. Recent advances in machine learning and image processing techniques have shown that machine learning models can identify features unrecognized by human experts and more precisely/accurately assess common measurements made in clinical practice.
