PREDIWEAR: A Wearable Sensor-Based Disease Prediction System a Comprehensive Survey

Abstract

Wearable sensor technology has emerged as a transformative paradigm in modern healthcare, enabling continuous and non-invasive monitoring of physiological parameters in real time. This work presents PrediWear, a conceptual framework that integrates wearable sensor data with machine learning techniques to support early and accurate disease prediction. The primary aim is to analyze existing methodologies and technological advancements that contribute to wearable sensor-based predictive healthcare systems. A comprehensive survey approach is adopted, reviewing relevant literature, datasets, sensing modalities, and computational techniques. Key sensing technologies considered include electrocardiography (ECG), photoplethysmography (PPG), galvanic skin response (GSR), inertial measurement units (IMU), and biochemical sensors. Advanced computational paradigms such as edge computing, federated learning, and explainable artificial intelligence (XAI) are examined for their role in improving system efficiency, scalability, and interpretability. The study identifies major challenges including data quality issues, sensor drift, inter-individual variability, and regulatory constraints that affect real-world deployment. The findings indicate that integrating robust sensing technologies with intelligent data processing can significantly enhance early diagnosis and preventive healthcare. In conclusion, PrediWear-type systems demonstrate strong potential for enabling personalized and remote patient monitoring, with future research directed toward improving reliability, clinical validation, and large-scale implementation.