The Implementation of Facial Symmetry Assessment Before and After Orthognathic Surgery Using Transfer Learning

In recent years, deep learning has been widely used in medical care. The implementation of deep learning technology in medical care makes the judgment of medical results more accurate and saves evaluation time. For example, the preoperative evaluation of orthognathic surgery requires a lot of time to observe the facial symmetry of the clinical patient. This paper, in cooperation with Chang Gung Memorial Hospital, studies to judge the degree of facial symmetry through migration learning. The system uses contour map as the training data, and construct the system to classify and score the symmetry of the face. Therefore, doctors can quickly and easily assess the status of clinical patients and save evaluation time. In this case, the contour map of the face from 71 patients provided by the doctor is classified into four equal parts according to the scores given by the doctors. Machine learning is used to extract sample features and a set of trained module. For user interface, the doctor uploads the image through the webpage, and the system outputs the classified result to provide a reference for the doctor so that the doctor can further understand the part of the patient that needs correction.