An intervertebral joint osteoarthritis image feature evaluation method based on deep learning

Abstract

The present application relates to the technical field of medical image auxiliary diagnosis, and proposes a kind of intervertebral joint osteoarthritis image feature evaluation method based on deep learning, can simultaneously identify five kinds of FJOA image features such as joint space stenosis, osteophyte, hypertrophy, subchondral bone erosion and subchondral cyst, improve the comprehensiveness and efficiency of evaluation.The nnU-Net model is used to carry out high-precision segmentation on the intervertebral joint region, effectively improve the positioning accuracy;By introducing the shared feature extraction network based on ResNet-18 and five parallel classification sub-networks, the joint modeling of multi-scale semantic information is realized, and the recognition accuracy and generalization ability of the model are improved;Combined with Grad-CAM technology to generate activation heat map, the model interpretation basis is superimposed to the original image, the visual display of evaluation result is realized, the explainability and clinical applicability of model are enhanced, the FJOA image features in axial lumbar CT image from two center data sets can be quantitatively evaluated, to comprehensively quantify individual FJOA image features.

Images