Monitoring Heart Rate Recovery Using Wearable ECG Devices to Identify Health Risk Groups

Cardiovascular fitness and autonomic balance play crucial roles in determining long-term health outcomes. This project focuses on developing a continuous heart health monitoring system utilizing ECG signals captured from wearable devices to analyze Heart Rate Recovery (HRR) dynamics. HRR, defined by the speed at which the heart rate recovers post-exercise, is an essential indicator of cardiovascular health. The system aims to classify individuals based on HRR patterns into high-risk (slow HRR) and low-risk (fast HRR) categories, facilitating early detection of potential cardiovascular risks.

The dataset comprises multiple CSV files containing time-domain, frequency-domain, and nonlinear features derived from ECG data. Key features include Mean RR intervals, SDRR, RMSSD, LF/HF ratios, and nonlinear metrics such as Sampen and Higuchi dimension. These features are employed to predict heart rate values using regression techniques and assess HRR dynamics.

The primary objective is to predict instantaneous heart rate (HR) from ECG-derived features, assess HRR patterns, and classify individuals based on their heart rate recovery speed. This real-time health monitoring system will help individuals identify and monitor their cardiovascular health through wearable devices, ultimately contributing to personalized healthcare interventions. The project also explores the potential to distinguish high-risk individuals based on the recovery rate after exercise.

Keywords: Heart Rate Recovery (HRR), ECG signals, Wearable devices, Heart Rate Prediction, Regression models, Time-domain features, Frequency-domain features, Nonlinear features, Cardiovascular fitness, Risk classification, Health monitoring.