Lenke Classification of Scoliosis Based on Segmentation Network and Adaptive Shape Descriptor

Scoliosis is a common spinal deformity that seriously affects patients’ physical and mental health. An accurate Lenke classification greatly significant for evaluating treating scoliosis. Currently, the clinical diagnosis mainly relies on manual measurement; however, using computer vision assists with an intelligent diagnosis. Due to complex rules of characteristics medical imaging, fully automated scoliosis remains considerable challenge. Herein, novel method X-rays based segmentation networks adaptive shape descriptors proposed. Three aspects our should be noted in comparison previous approaches. We used Unet++ segment vertebrae designed post-processing operation improve effect. Then, we proposed new descriptor extract features segmented greater detail. Finally, framework contains two schemes Cobb angle measurement classification, respectively. After rigorous experimental evaluations public dataset, achieved best performance outperformed other sophisticated