AI‐Assisted Self‐Powered Wearable Dual‐Mode Sensor With TENG and Stretchable Optical Fiber for Neurological Disorder Diagnostics

Wearable sensors hold significant potential for managing lower-limb dysfunction in neurological disorders, but current systems remain constrained by unimodal sensing, external power dependence, and limited diagnostic capabilities. Here, we present a wireless wearable dual-mode sensor (WDMS) integrating three polyurethane-based flexible optical strain (PFOS) components with a contact-separation mode triboelectric nanogenerator (CS-TENG). The PFOS components are used for muscle signal monitoring, while the CS-TENG simultaneously monitors plantar pressure and harvests biomechanical energy to power the WDMS, eliminating external power dependence. Furthermore, leveraging gait data from 60 individuals, an embedded Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model achieved 94.23% accuracy in distinguishing Parkinson's disease (PD) and stroke, while quantitatively evaluating rehabilitation progress after pharmacological and physical therapy interventions. By synergizing multimodal sensing, AI-driven analysis, and clinical validation, this technology advances beyond passive monitoring to provide intelligent diagnostic support. Its self-sufficiency and scalability facilitate transformative home-based management of neurological disorders.