A Scalable AI Framework for Proactive Health Monitoring Using Physiological Signals

This paper presents a scalable framework for early health monitoring in diverse care settings, leveraging four key vital signs—heart rate, respiratory rate, temperature, and oxygen saturation—that can be captured through commonly available means or unobtrusive wearable sensors. The framework integrates clinical criteria with advanced AI techniques, including ensemble modeling and an adapted BERT-based clustering approach, to classify individuals into health risk categories for timely intervention. Validated on publicly available PhysioNet CinC and MIMIC-III datasets, the framework demonstrates consistent and robust performance. The ensemble model achieves high precision and recall for non-sepsis detection (F1 score:0.99), with moderate performance for sepsis classification, while the BERT model exhibits strong discriminative ability, achieving ROC-AUC values of 0.83 and 0.90, and PR-AUC values of 0.49 and 0.89 on PhysioNet and MIMIC-III, respectively. These results underscore the framework’s adaptability and effectiveness across diverse data distributions. We also present a real-time interface that provides actionable insights for caregivers, enabling timely interventions and addressing resource constraints.