Backpropagation Neural Network Artificial Intelligence Algorithm-Based Magnetic Resonance Imaging Image Feature Analysis in the General Anesthesia Hip Arthroplasty

Abstract

Objective. This study aimed to present an investigation of the clinical significance of magnetic resonance imaging (MRI) images obtained based on the backpropagation neural network (BPNN) artificial intelligence algorithm for hip arthroplasty under general anesthesia. Methods. In this study, a case-review method was used to collect 100 patients requiring total hip replacement. They were then randomly divided into an observation group and a control group. Based on the neural network algorithm, the images of the two groups of patients were analyzed to judge their accuracy. Then the sensitivity, specificity, and accuracy of MRI images based on neural algorithms were compared with those processed by radiologists. Results. It was found that MRI processed by BP neural network had good accuracy in the diagnosis of hip joint diseases compared with CT. Meanwhile, the images processed by BP neural network had good specificity and accuracy compared with the images processed by radiologists. Conclusion. Imaging images obtained by BPNN artificial intelligence algorithm were more accurate than CT images, which had more guiding value for surgeons in operation.