Precision at Heart: An IoT-based Vertical Federated Learning Approach for Heterogeneous Data-Driven Cardiovascular Disease Risk Prediction

Abstract

<p>Cardiovascular disease (CVD) encompasses a wide range of diseases that affect the heart and blood vessels, including coronary artery disease, heart failure, arrhythmia, and stroke. Machine Learning (ML) has been widely used to predict CVD risk based on various factors and is a critical area of healthcare research. However, due to privacy concerns, sharing the data needed to predict CVD with ML is challenging. Even though Federated Learning (FL) enables distributed training of ML models without sharing raw data, it assumes that all training features are available to all clients. To address the problem, we propose a Vertical Federated Learning (VFL) based method that trains ML models in a distributed manner and has different features available to several spatial locations. In this work, each party maintains a portion of separate data features, performs calculations on them locally, and then transfers only the necessary information to jointly train an ML model. We employ the proposed method for different use cases where the data characteristics are split between: i) the patient and the hospital (2 splits); ii) the patient, the doctor, and the laboratory (3 splits); and iii) the patient, the doctor, the Electrocardiogram (ECG) center, and the laboratory (4 splits). We test the proposed methodology on the realistic publicly available dataset</p>