This invention provides a method and
system for predicting the risk of preoperative condylar
bone resorption, relating to the field of
oral medicine technology. The method includes the following steps: acquiring and preprocessing small-field
CBCT image data of the
temporomandibular joint region of the patient to obtain the corresponding
region of interest (ROI) of the condylar
cancellous bone region; segmenting and extracting features from the ROI of the condylar
cancellous bone region to obtain corresponding
radiomics features; constructing a condylar
bone resorption risk prediction model and inputting the
radiomics features into the model for prediction to obtain the corresponding risk prediction results; generating
patient risk grading information based on the risk prediction results to assist clinicians in making
patient management decisions. This invention, by analyzing and modeling the
radiomics features of the patient's preoperative condylar
cancellous bone region, achieves the prediction of the risk of postoperative condylar
bone resorption without using any postoperative images as prediction input.