Development of a wearable sensor system for real-time cardiovascular monitoring

Thesis Abstract

Cardiovascular diseases remain a leading cause of mortality worldwide, necessitating continuous and real-time monitoring of vital cardiac parameters to enable early detection and prompt intervention. This study addresses the critical need for accessible, reliable, and non-invasive solutions for ongoing cardiovascular assessment by developing a wearable sensor system capable of delivering real-time data on key indicators such as heart rate, blood oxygen saturation, and electrocardiogram (ECG) signals. The primary aim of this research is to design, implement, and validate an innovative wearable device that facilitates continuous cardiovascular monitoring outside clinical settings, thereby promoting proactive health management. The specific objectives include (1) identifying and integrating appropriate biosensors for cardiac parameter detection, (2) developing a low-power and user-friendly wearable prototype, (3) establishing wireless data transmission protocols, and (4) evaluating system accuracy and reliability through clinical validation. A mixed-methods research design was adopted to achieve these objectives. Quantitative data collection involved constructing a prototype wearable sensor system integrated with photoplethysmography (PPG), dry ECG electrodes, and pulse oximetry sensors, connected via Bluetooth Low Energy (BLE) to a smartphone application. The population targeted included 150 adult volunteers aged 20-65 years, comprising both healthy individuals and patients with diagnosed cardiovascular conditions, recruited through hospital partnerships. A stratified sampling technique was employed to ensure demographic and health status diversity. Data collection instruments encompassed the sensor hardware, a custom-developed mobile application, and validated clinical measurement devices for benchmarking purposes. The validity and reliability of the sensor data were established through calibration against gold-standard clinical equipment, employing Bland-Altman analysis and intraclass correlation coefficients (ICC). Data analysis involved descriptive statistics to summarize baseline characteristics, followed by inferential statistical methods, including regression analysis to assess the correlation between wearable sensor readings and clinical standards, and machine learning algorithms, such as support vector machines (SVM), for anomaly detection. The system's performance was further evaluated using receiver operating characteristic (ROC) curves to determine sensitivity and specificity. Expected findings indicate that the integrated wearable sensor system will demonstrate high accuracy, with ICC values exceeding 0.9 in comparison to clinical gold standards, and ROC-AUC scores above 0.85 for atrial fibrillation and other arrhythmia detection. The study anticipates revealing key relationships between sensor-derived data and clinical assessments, confirming the system's potential for reliable real-time monitoring. The research contributes to the growing body of knowledge by providing a validated model for portable cardiovascular surveillance that combines sensor technology with wireless communication and data analytics, aligning with the Health Belief Model and the Technology Acceptance Model to inform user adoption strategies. The study concludes that the developed wearable sensor system offers a viable solution for continuous cardiovascular monitoring outside clinical environments, with significant implications for early diagnosis, risk stratification, and remote patient management. Based on these findings, recommendations include further scaling of the prototype for long-term field trials, integration with electronic health records, and exploration of artificial intelligence-driven predictive analytics. Future research avenues should investigate customization for specific populations, such as elderly users or athletes, and evaluate long-term adherence and usability to maximize clinical impact.