AI-Enhanced Smart Textile System for the Monitoring of Cardiovascular Insufficiency

This study explores the integration of artificial intelligence (AI) and sensor fusion in wearable technology for early cardiac emergency detection and improved non-invasive cardiac diagnostics. Developed as part of a doctoral thesis within the Maia project in collaboration with Fraunhofer IZM and Charité Berlin, the system combines bioimpedance spectroscopy (BIS) with complementary sensor technologies, enabling synchronous monitoring of over 110 cardiac parameters at high data rates. Clinically relevant parameters were selected with input from Charité clinicians, ensuring precision. The system employs biocompatible silicone-based sensors and optimized semi-dry textile electrodes embedded in a smart textile vest, allowing continuous, non-invasive cardiac monitoring. By integrating BIS with electrocardiography (ECG), seismocardiography (SCG), phonocardiography (PCG), and photoplethysmography (PPG), this comprehensive system supports in-depth cardiac health assessment. Advanced deep learning and statistical machine learning algorithms enhance diagnostic accuracy, with models trained on clinical and patient-specific data, showing high sensitivity, specificity, and accuracy in detecting cardiovascular conditions. Validation with data from the PhysioNet database underpins this foundational design and development study.