Spine CT Image Segmentation Algorithm Based on Multi-scale Residual Network
In order to improve the segmentation accuracy of spine computed tomography(CT)images,this paper proposes an optimized segmentation method based on the U-Net network.In the feature extraction stage,the multi-scale residual network is used to optimize the structure and improve the quality of image feature extraction.At the same time,an attention module is used in the up-sampling process to optimize the extraction of bone edge information.Experimental results show that the Dice coefficient,accuracy rate,precision and recall rate of the algorithm reach 95.82％,99.62％,96.05％and 99.83％respectively,which are higher than similar algorithms,and effectively improve the segmentation quality of medical CT images.
万方数据
维普
