基于3D体素增强和3D alpha背景分离的多发性硬化症病灶分割方法-《中国医学物理学杂志》

文章信息/Info

Title:: Multiple sclerosis lesions segmentation based on 3D voxel enhancement and 3D alpha matting

作者:: 孙颖1; 张吟龙2; 王鑫3; 曾子铭4; 毛海霞4; 1.中国医科大学附属第一医院重症医学科，辽宁沈阳 110001； 2.中国科学院沈阳自动化研究所，辽宁沈阳 110016； 3.沈阳建筑大学信息与控制工程学院，辽宁沈阳 110168； 4.深圳职业技术学院汽车与交通学院，广东深圳 518055

Author(s):: SUN Ying1; ZHANG Yinlong2; WANG Xin3; ZENG Ziming4; MAO Haixia4; 1. Department of Critical Care, the First Hospital of China Medical University, Shenyang 110001, China 2. Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, China 3. College of Information and Control Engineering, Shenyang Jianzhu University, Shenyang 110168, China 4. School of Automotive and Transportation Engineering, Shenzhen Polytechnic, Shenzhen 518055, China

Keywords:: Keywords: multiple sclerosis lesion segmentation 3D voxel enhancement 3D alpha matting color image segmentation technique

摘要:: 目的：提出一种用于T1加权像、T2加权像和流体衰减反演恢复（Flair）磁共振图像的多发性硬化症（MS）病变分割方法。方法：首先基于3D图像增强技术，将高强度MS病变区域与其他组织区域区分开来。然后利用假阳性降低方法，去除一些强度和密度不均匀的假阳性目标区域（VOI），并利用颜色分割法去除白质之外的VOI。最后利用彩色MR技术生成3个区域，以便细化分割MS病变。结果：在CHB数据集上进行测试，得到真阳率均值为0.48，Dice相似系数均值为0.52。结论：该方法能够有效去除噪声及其他无关非病变组织，并能准确识别并分割MS病变，该方法的有效性、准确性能为后续的MS分割技术分析提供依据。同时为MS病变的预防治疗、病情跟踪提供客观、方便的诊疗方法。【关键词】多发性硬化症；病灶分割；3D体素增强；3D alpha背景分离；颜色分割技术

Abstract:: Abstract: Objective To propose a novel method for multiple sclerosis (MS) lesions segmentation in T1-weighted, T2-weighted and fluid-attenuated inversion recovery (Flair) MRI images. Methods MS lesions with high intensity were distinguished from other tissues using 3D image enhancement technology. Then, the false positive volume of interest (VOI) with uneven intensity and density were removed by the false positive reduction method, and the VOI outside the white matter were eliminated by the color image segmentation method. Finally, the color MR technique was used to generate 3 regions to refine MS lesions segmentation. Results The test on CHB dataset showed that the mean true positive rate and mean Dice similarity coefficient reached 0.48 and 0.52. Conclusion The proposed method can not only remove noise and other non-pathological tissues effectively, but also identify and segment MS lesions accurately. Because of its effectiveness and accuracy, the proposed method can provide a basis for the subsequent analysis of MS segmentation technology, and provide on objective and convenient method for the prevention and treatment of MS lesions and disease tracking.