北京生物医学工程

基于皱褶轮廓贝塞尔曲线插值的肠道内镜图像三维重建

Colon 3D reconstruction from endoscope images based on colonfold Bezier curve interpolation

作者：施展廖洪恩

单位：清华大学医学院生物医学工程系 (北京100084)

分类号：R318.04

出版年·卷·期（页码）：2016·35·6（576-581）

摘要：

目的肠道癌症的及早发现与治疗能大大提高患者的存活率，肠道内镜图像的三维重建可以增强医生对肠道结构的理解。目前肠道三维结构重建方法受肠道皱褶遮挡影响无法重建完整的肠道结构，并且重建的点云不能直观呈现肠道的结构。本文提出了基于肠道褶皱插值的重建方法拟解决重建的肠道不完整以及点云显示不直观的问题。方法首先利用肠道呈管道状的先验信息，运用运动恢复结构技术三维重建肠道褶皱轮廓，然后在皱褶之间进行贝塞尔插值重建肠道褶皱之间的肠壁，实现肠道整体的三维结构重建。最后在OpenGL平台上，渲染并显示肠道的三维结构。本文应用该方法从两张内镜肠道图像重建了肠道的结构。结果该方法重建的肠道结构轮廓位置准确，能清晰直观地展现肠道结构。该方法融入了肠道形状先验信息，从内镜图像重建并直观显示了肠道三维结构。结论基于皱褶轮廓贝塞尔曲线插值的肠道内镜图像三维重建方法，克服了传统三维重建的缺点，能为医生提供直观的肠道结构信息，也为从连续多帧内镜图像重建肠道结构提供了技术基础。

Objective Early detection and treatment of colon cancer can raise the patient survival rate greatly, and reconstruction of endoscopic colon images can improve doctor’s understanding of the structure of colon. We propose a method to reconstruct the colon structure to overcome the weakness of existing methods. Methods The proposed method takes advantage of the prior information that the shape of the colon is tubular. We reconstruct the colon fold by using structure from motion technique, and interpolate the surface wall between the colon fold by using Bezier curve. Finally, we render and display the 3D colon structure with OpenGL. Results The results of the proposed method reconstruct accurate colon fold position and show the colon structure clearly. Conclusions The reconstruction method overcomes the shortcomings of traditional 3D reconstruction, provides intuitional structure information for the doctors, and is critical toward virtual endoscopy of colon from colonoscopy videos.

参考文献：

［1］李景南，赵莉，苗政.结肠癌和直肠癌发病模式和临床特点的异同分析［J］.中国实用内科杂志，2010，30（10）：916-918

Li Jingnan，Zhao Li，Miao Zheng.The difference of morbidity and clinical characteristics betweens colon cancer and rectal cancer［J］.Chinese Journal of Practical Internal Medicine，2010，30（10）：916-918.

［2］张瑜，王磊，郭继中.结肠癌的早期诊断［J］ .医学综述，2009，15（20）：3083-3086.

Zhang Yu，Wang Lei，Guo Jizhong. The early diagnosis of colon cancer［J］. Medical Recapitulate，2009，15（20）：3083-3086.

［3］Hu M, Penney G, Edwards P. 3D reconstruction of internal organ surfaces for minimal invasive surgery ［J］.MICCAI 2007 Lecture Notes in Computer Science, 2007, 4791: 68-77.

［4］Hartley R,Zisserman. Multiple View Geometry in Computer Vision Second Edition［M］.UK: Cambridge University Press,2003:239-261.

［5］Bouguet JY. Camera Calibration Toolbox for MATLAB［CP/DK］. http://www.vision.caltech.edu/bouguetj/calib_doc/index.html.

［6］Bay H，Tuytelaars T, Gool LV. SURF: speeded up robust features［M］//Leonardis A, Bischof H, Pinz A. ECCV 2006 Lecture Notes in Computer Science. Berlin Heidelberg: Springer, 2006: 404-417.

［7］Fischler MA，Bolles RC. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography［J］.Communications of the ACM, 1981,24(6):381-395.

［8］Canny JF. A computational approach to edge detection［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986,8(6): 679-698.

服务与反馈：

【文章下载】【加入收藏】

提示：您还未登录，请登录！点此登录