Pallak Dhabalia
Research Mentor(s): Mia Bonini
Mentor Department: Biomedical Engineering
Authors: Pallak Dhabalia, Mia Bonini, Feng Gu, Dan Beard
Session: Session 4 (1:00pm – 1:50pm)
Presentation Type: Poster 103
Abstract
Left Ventricular Assist Devices (LVADs) are life-sustaining mechanical pumps used for patients with severe heart failure. Despite their benefits, patient outcomes following LVAD implantation vary significantly, necessitating improved predictive models for personalized treatment planning. This research aims to enhance outcome prediction in LVAD patients by leveraging computational modeling and machine learning techniques. Our approach involves modifying a 0D computational model to simulate cardiovascular dynamics in LVAD-supported patients. By integrating patient-specific clinical data, we develop a phenomapping framework that identifies subgroups with similar hemodynamic characteristics. Additionally, machine learning algorithms are applied to analyze these subgroups and predict post-implantation outcomes, including survival rates, complications, and quality of life measures. This research has the potential to refine clinical decision-making by providing more accurate prognostic tools for physicians. By combining computational simulations with data-driven methods, we aim to contribute to the advancement of precision medicine in heart failure treatment. Future work will focus on validating the model with larger datasets and exploring real-time predictive applications.
