Isotonic Muscle Strength Evaluation Based on Ultrasound Video

In order to improve the simplicity and accuracy of detecting muscle strength, a recognition method based on deep learning is proposed. CNN algorithm and RNN algorithm are used to complete the classification of B-ultrasound images and videos with different muscle strength. The convolutional neural network CNN algorithm is used to extract the features of B-ultrasound video, and then the cyclic neural network RNN is used to complete the time sequence analysis. The experiment uses the data set in this paper for training and testing. The results show that the accuracy of the proposed model for four categories of muscle strength video based on 0%, 20%, 40% and 60% of individual maximum strength (MVC) is about 93.45%. The resnet50 RNN algorithm proposed in this paper can effectively classify B-ultrasound videos with different muscle strength. This muscle strength detection model has strong universality for different individuals. The method of processing B-ultrasound image video can provide theoretical guidance for medical rehabilitation treatment, prosthetic limb control, professional sports training, public health and other fields, and promote the development of these fields, which is worthy of our in-depth study.