Abdul Aziz Jamal Eddin
Research Mentor: Mohammed Saeed
Mentor Department: Department of Internal Medicine, Medicine
Author(s): Abdul Aziz Jamal Eddin, Mohammed Saeed
Session: Session 7 (4:00 PM – 4:50 PM)
Presentation Type: Poster 78
Abstract
Patients with heart failure and bundle branch block on their ECGs often undergo treatment with cardiac resynchronization therapy (CRT). The goal of CRT is to improve heart function and reduce symptoms of heart failure. However, approximately 30% of patients who undergo CRT are non-responders based on post-operative assessment. Identifying patients who are likely to be non-responders prior to procedures would help to reduce risk associated with an invasive CRT procedure and unnecessary healthcare costs. We developed machine learning models to predict favorable response prior to surgical operation. Our model leverages a set of pre-operative patient data with known post-operative outcomes to train machine learning algorithms. We evaluated model accuracy in predicting response patterns in a test set of patients. We hypothesize that such ML techniques may help identify patients who are likely to be more responsive to CRT prior to procedures.
