A Wearable-Based Hidden Markov Model for Health Monitoring in Conflict Zones: A Case Study of Gaza

Abstract

During humanitarian crises, such as the ongoing genocide in Gaza, the collapse of healthcare infrastructure severely disrupts patient intake processes. Unconscious individuals and unaccompanied children arrive at medical facilities without any means of identifying their age, blood type, or medical history -- information that is critical for effective treatment. Severe shortages in medical personnel and essential diagnostic tools further hinder the ability to monitor vitals and track patient conditions during treatment. Current health monitoring wearables are expensive and often ill suited for crisis settings where power and resources are scarce. This thesis proposes the development of a rugged and cost-effective wearable device designed for use in conflict zones. The device integrates both passive and active components: a passive NFC module stores essential patient identification and medical history, allowing healthcare workers to access critical data efficiently, while active sensors monitor patient vitals. To extract meaningful insights from the data, Hidden Markov Models (HMMs) are employed to assess physiological states over time. These models account for time-dependent variations in patient vitality, providing a probabilistic framework to detect stress, fatigue, and critical health conditions. By incorporating age and gender specific physiological baselines, the HMMs enhance the device’s ability to adapt to individual health patterns, improving its effectiveness in triaging patients.