Cardiovascular Disease Risk Prediction For People Living With Hiv Using Ensemble Deep Neural Network

Abstract

The presence of cardiovascular disease (CVD) alongside Human Immunodeficiency Virus (HIV) presents a significant challenge in public health, with HIV-positive individuals experiencing a heightened risk of CVD-related morbidity and mortality. Despite advancements in antiretroviral therapy (ART), people living with HIV (PLWH) still face elevated susceptibility to CVD, necessitating tailored approaches to risk prediction and management. Existing CVD risk prediction models often overlook HIV-specific variables, leading to inaccurate assessments of CVD risk in PLWH. This research aims to address this gap by developing an Ensemble Machine Learning (ML) model to increase the accuracy of CVD risk prediction by combining conventional risk factors with HIV-specific characteristics among PLWH. The study involves a thorough analysis of the collection of studies on CVD prediction using machine learning techniques and the compilation of clinical datasets from diverse global regions, including Sub-Saharan Africa and Asia. Unconventional risk factors such as ART, chronic inflammation, and genetic markers are incorporated into the ML models to improve predictive accuracy. The proposed model serves as a clinical decision support tool and offers individualised risk assessments, empowering medical practitioners to make well-informed choices and improve HIV treatment strategies for individuals living with the virus. This work has the potential to revolutionise the prediction and management of cardiovascular disease risk by utilising advanced machine learning techniques like deep learning. This would ultimately lead to better health outcomes and a higher standard of living for people living with HIV worldwide.