Nanjing University of Chinese Medicine Researchers Provide New Data on Orthopedic Surgery (Supraspinatus extraction from MRI based on attention-dense spatial pyramid UNet network).

Researchers from Nanjing University of Chinese Medicine have developed a deep learning model to automatically extract the supraspinatus, a common site of rotator cuff tears, from MRI images. The model, called A-DAsppUnet network, was trained and optimized using a dataset of 60 patients with rotator cuff tears. The results showed that the proposed model had a precision of 99.20%, an intersection over union of 83.38%, and a Dice coefficient of 90.94%. The researchers concluded that the model accurately extracts the supraspinatus from MRI and holds practical significance in assisting medical decision-making. [Extracted from the article]