Paper
One of the surgical devices used to treat cervical spine disorders. The atlantoaxial lateral block is fusion device. Atlantoaxial joint space reconstruction is one of the key steps in the use of the atlantoaxial lateral block fusion device, whereas the conventional 3D atlantoaxial joint space reconstruction suffered from low reconstruction precision and accuracy, as well as the inability to take into account its dynamic properties accurately. To address these issues, this work proposes a parallel segmentation reconstruction model. By using the patient's cervical spine CT datas as input, the atlantoaxial joint gap is reconstructed in 3D by the gap edge detection module and 3D reconstruction module of the model in this paper, and the visualized 3D model is output. In the gap edge detection module, an advanced image segmentation algorithm based on Cxy-Net is adopted to optimize and extract the details of the gap. The average Hausdorff distance (Hd) of this model is 10.5211 mm, the average symmetric surface distance (ASD) is 0.3861 mm, the average surface overlap (So) reaches 90.09%, the average Dice similary coefficient (Dice) is 0.8834, and the average accuracy (AC) is 0.8914. Compared with the conventional modeling, the model of the present paper improves the accuracy, Dice similary coefficient, and accuracy by about 15.37%, 8.96%, and 4.84% respectively.