Muscle fatigue state classification based on blood flow bioimpedance

The development of physiological research has established the importance of muscle fatigue assessment in all aspects of medical research, ergonomics, sports injuries, and human-computer interaction. The fatigue of muscle during intermittent exercise is closely related to blood flow. Bioelectrical impedance technology can be used to describe blood flow and reflect fatigue. In this paper, we propose to classify the blood flow impedance signals of human muscle fatigue through machine learning, so as to achieve the purpose of nondestructive testing and evaluation of muscle fatigue. First acquisition in the process of fatigue blood flow impedance signal, the signal filtering denoising processing and extraction of characteristic value, and then build a gcForest model after preprocessing of the data is divided into three categories: the fatigue state, transition fatigue state, and the depth of the fatigue state, and the traditional machine learning method of support vector machine (SVM) model and the neural network model were compared. The results show that the overall fatigue recognition accuracy and evaluation indexes of the gcForest model are better than those of the traditional methods, and the overall accuracy of the gcForest model is 92%. This method provides a new technique for muscle fatigue state detection.