Rib Fracture Segmentation Based on 3D CE-Unet

Due to the structural complexity of the rib cage and the diversity of fracture forms, traditional image segmentation methods often struggle to cope, resulting in unsatisfactory performance in terms of accuracy and robustness. These limitations not only affect the accurate diagnosis of rib fractures, but also create barriers to treatment planning. To address this challenge, this paper proposes a new rib fracture segmentation method 3D CE-UNet, which is based on 3D U-Net network architecture and incorporates ECA (Efficient Channel Attention) and CBAM (Convolutional Block Attention Module) attention mechanisms. This combination not only improves the model's ability to recognize rib fracture features, but also enhances the capture of details in complex medical images, thus achieving higher segmentation accuracy.