A Strain Sensing-Based Diagnostic System For Temporomandibular Disorders

This study aims to develop a strain sensing-based hydrogel electronic sensor (PVA-EBPD) for the intelligent diagnosis of temporomandibular disorders (TMD), addressing the limitations of traditional diagnostic methods such as high physician dependency, complexity, cost, invasiveness, and radiation exposure. The performance of the PVA-EBPD hydrogel sensor was validated through 3D facial morphology analysis, dynamic covalent bond crosslinking hydrogel preparation, rheological testing, strain sensing performance evaluation, and biocompatibility experiments. In clinical trials, 120 volunteers participated in sensor testing, and the data were analyzed using continuous wavelet transform and long short-term memory network. In clinical trials, the PVA-EBPD hydrogel sensor achieved a diagnostic accuracy of 90.00% and a sensitivity of 100% for DDwR, with the predictive model showing an accuracy of 84.4% and a sensitivity of 86%. The sensor precisely captured minute strains during TMJ activities and identified the specific timing of abnormal joint movements, particularly demonstrating a sensitivity of 98.28% for joint clicking during mouth opening and closing. Additionally, the sensor exhibited instantaneous self-recovery within 2.1 seconds, an ultra-low detection limit of 0.005% strain, and strong skin adhesion with an adhesion force of up to 40 N, while maintaining excellent biocompatibility. Furthermore, the sensor successfully monitored TMJ activities, such as chewing, in both humans and pigs. The sensor with its instantaneous self-recovery, ultra-low detection limit, favorable skin adhesion, and excellent biocompatibility, provides a novel, non-invasive, and efficient tool for TMD diagnosis, demonstrating significant potential for clinical applications.