Development of a Biomechanical Diagnosis and Analysis System Using a Textile Elbow Angle Sensor: Integrating Inverse Dynamics and Multi-Layer Perceptron Techniques

The recent development of algorithms through artificial intelligence and the ability to measure the human body through soft textile sensors has enabled the provision of meaningful information to the wearer. In this study, a sensor sleeve using a textile elbow angle sensor that can measure the bending and relaxation of the elbow was manufactured and measured. In addition, biomechanical data from Biomechanical of Bodies (BoB)-4, a software capable of inverse dynamics that can optimally calculate the load on human joints and segments during exercise, was collected. A continuous system of resistance angle and angle biomechanical data was designed with an artificial intelligence multilayer perceptron (MLP) algorithm, and the accuracy and output results were checked. Consequently, the accuracy of MLP1 and MLP2 is exceedingly high, at approximately 0.80 and 1.00, respectively. The biomechanical data of the system is comparable to that of BoB, rendering it suitable for providing reliable information to the wearer. Based on this study, it is possible to develop algorithms and systems that can perform biomechanical analysis for various exercise movements in the future.