Wearable Multimodal Sensing System for Synchronously Health–Environmental Monitoring via Hybrid Neuroevolutionary Signal Decoupling

Abstract

Wearable sensors are advancing personalized healthcare and environmental safety, yet integrating multimodal sensing remains challenging due to material incompatibility, cross-sensitivity, and environmental interference. To address this gap, we present a monolithic wristband-integrated multimodal sensing platform for simultaneous environmental and physiological monitoring. The system integrates sensors for NO₂, UV irradiance, epidermal temperature, and human pulse signals. A TiO₂/WS₂ heterojunction-based resistive transduction matrix, as the core sensing material, achieves room-temperature NO₂ detection (theoretical limit of detection of 14.4 ppb) with 12-month stability, UV intensity measurement (0.024-1.68 mW/cm²), epidermal temperature monitoring (25-50 °C, sensitivity of 0.22%/°C), and arterial pulse waveform analysis (P-T-D peak resolution). A hybrid neuroevolutionary algorithm (GA-BP) decouples photo-gas interference, reducing quantification errors to <3.5%. Flexible electronics and selective encapsulation (PDMS, PET shielding) ensure mechanical durability and accurate signal acquisition. The platform demonstrates multifunctional capabilities that enable point-of-care health and environmental monitoring, bridging personalized diagnostics with exposure assessment.