Machine Learning Algorithm to Guide Cardiac Resynchronization Therapy Personalization

Technical Summary:

The patient-response assessment machine learning algorithm is a good predictor of patient benefit, or lack thereof, from cardiac resynchronization therapy. This algorithm utilizes patient characteristics including but not limited to: demographics, comorbidities, medication history, circulating protein biomarkers, and imaging data. A clearer picture of patient response to CRT is generated that, when coupled with existing methods to inform patient candidacy, can improve outcomes. These outcomes include both increased intervention success rates and reduced unnecessary adverse event occurrences. This algorithm has been proven to correctly predict patient outcomes in previously conducted CRT clinical trials.