Study on automatic localization algorithm of Baliao points

In the clinical treatment of modern traditional Chinese medicine, such as reproductive system and gynecological diseases, acupuncture at Baliao points has become a primary method. An improved YOLOv7 model is proposed to train and predict the CT images of Baliao points, which will automatically detect the sacral cornu and eight acupoints, as well as calculate the required distances and angles. The experimental results show that the detection accuracy and speed are significantly improved. The mean average precision and FPS have reached 93.1% and 78.9, respectively, which are 4.6% and 25.5 higher than those before the improvement. By comparing these predicted values with the actual values provided by the hospital, the results show that the predicted values are within the safe error range, which will provide data support for subsequent research on exogenous auxiliary devices.