IoT system development for heart rhythm monitoring and cardiovascular risk estimation

Abstract

The research focuses on addressing the global issue of cardiovascular diseases. The key variable under consideration for predicting cardiovascular diseases is heart rate variability (HRV). Leveraging the widespread adoption of IoT in various applications, particularly in the health sector, the study proposes the design and implementation of an IoT system for HRV monitoring. The research unfolded in four methodological phases: exploration and selection of technologies, definition of the IoT architecture, development of the prototype, and verification of its functionality. The implemented IoT system adheres to the conventional 4-layer IoT architecture: capture, storage, analysis, and visualization. Heart rate data is periodically acquired using a heart rate sensor and an Arduino-compatible board. The storage layer employs a non-relational database to store the captured data. The analysis layer extracts metrics related to HRV (High: RR <750 ms, Moderate: RR 750–900 ms, Low: RR >900 ms) by applying and delivering quantitative results from clustering algorithms such as machine learning models to evaluate data distribution. Risk levels indicate specific patient metrics. Thus, a 75-year-old patient exhibits an average HR of 75.56, Avg. RR of 795.42, falling into Cluster 1 with a risk value of 1.0. Similar detailed metrics and risk stratifications are presented for patients aged 68, 46, 37, and 18, demonstrating the system's robustness and efficacy in assessing cardiovascular risk. The visualization layer enables real-time observation of physiological variables, risk metrics, and results from data analytics models. The distinctive features of the results lie in the portability advantages of the IoT system, utilizing free hardware and software tools. This facilitates easy replication and utilization of the proposed system in medical campaigns, specifically for the early detection of cardiac conditions. The portable IoT system, leveraging free tools, enhances predictive capabilities for early cardiovascular risk detection globally