1、生命科学与技术学院研究生课程简介课程名称：现代医学图像处理 课程代码：170.494课程类型： 博士专修课程 硕士专修课程考核方式：全英文考试 教学方式：全英文讲授适用专业：生物医学工程 适用层次： 硕士 博士开课学期：秋季 总学时：32 学时 学分：2先修课程要求：医学图像处理课程组教师姓名 职 称 专 业 年 龄 学术方向丁明跃（负责人） 教授 生物医学工程51 医学图像处理，超声成像张旭明 副教授 生物医学工程37 医学图像处理，超声成像侯文广 讲师 生物医学工程38 医学图像处理，计算机视觉课程负责教师留学经历及学术专长简介：丁明跃，男，51 岁。华中科技大学生命科学与技术学院院长助理
2、、生物医学工程系常务副主任，教授、博士生导师。1991 年 9月-1993 年 7月，德国不伦瑞克工业大学博士后，德国洪堡基金洪堡学者；1996 年 12月-1997 年 2月，1999年 7-9月德国不伦瑞克工业大学机器人研究所高级访问学者；2000 年 5月-7月香港城市大学机械制造与管理工程系研究员；2000 年 10月-2005 年 4月，加拿大 Robarts研究所研究员。在此期间，分别为 Robarts研究所举办了 10个学时的序列学术讲座，并且在 2004年秋季为加拿大西安大略大学研究生讲授了26学时的现代医学图像处理课程（课程代码 BME519。自 1985年以来，一直从事图像
3、处理相关研究和教学工作。2005 年回国以后，专门从事医学图像处理相关研究与教学工作，共承担完成国家自然科学基金，国家 973、高技术863，国家支撑计划，教委博士点专项基金等三十余项。已编辑出版中英著作八部，在国内外权威期刊以及国际会议上发表论文 100余篇，其中被 SCI 收录 15篇，EI 收录 100篇，ISTP 收录 50篇。获欧洲专利一项，中国发明专利四项，计算机著作登记权一项，申请美国专利三项，国内发明专利 6项。先后获得德国洪堡基金，加拿大青年创新奖等国内外奖励 23项。是 IEEE高级会员，SPIE会员，中国电子学会、中国生物医学工程学会高级会员， 电子学报 宇航学报编委，中
4、国医疗器械技术创新联盟专家委员会委员，湖北省生物医学工程学会生物医学仪器专业委员会副主任委员，中国生物医药生物技术协会生物医学信息分会常务委员，中国电子学会生命电子学分会委员，中国机械工程学会再制造分会委员。课程教学目标：介绍医学图像处理处理领域常用的图像分割和图像配准方法；提供学生英语听说机会，掌握常用的术语和概念；通过全英语课程、提问与回答、考试等教学环节，提高学生英文科技阅读能力和运用能力。课程大纲：第一章 简介（两个学时）1.1 什么是医学图像处理？1.2 图像处理与图像分析的区别1.3 医学图像处理中最重要的两项任务：分割与配准1.4 课程考试构成第二章 基于神经网络的图像分割方法（
5、四个学时）2.1 什么是人工神经网络？2.2 基于神经网络图像分割的基本思想 2.3 采用自组织神经网络聚类的图像分割方法2.4 采用纹理特征的自组织神经网络图像分割方法第三章 可变模型方法（四个学时）3.1 谁发明了可变模型方法？3.2 什么是可变模型？3.3 轮廓的表达3.4 内力与外力3.5 轮廓的变形3.6 可变模型方法的发展第四章 三维超声图像分割（四个学时）4.1 三维超声成像4.2 前列腺近放射治疗4.3 三维超声图像针分割方法4.4 三维超声图像前列腺分割方法4.5 三维超声图像放射性粒子分割方法第五章 图像配准的基本概念（四个学时）5.1 什么是图像配准？5.2 什么要进行图
6、像配准？5.3 图像配准基本过程5.4 图像配准中的几何变换5.4 图像配准方法的分类第六章 基于粗精搜索的两级图像配准方法（四个学时）6.1 两级配准框架结构6.2 基于特征的图像粗配准6.3 基于区域的图像精配准6.4 基于 Harris角点检测的图像配准方法第七章 采用互信息的图像配准方法（六个学时）7.1 信息的度量7.2 熵的定义7.3 什么是互信息？7.4 互信息的性质7.5 基于互信息图像配额准方法的框架结构7.6 基于最大互信息的 MRI与 PET图像配准方法第八章 非刚体图像配准（四个学时）8.1 刚体与非刚体8.2 非刚体变换在临床中的应用8.3 全局变换与局部变换8.4
7、相似性度量8.5 采用自由变形的非刚体图像配准及其在乳腺 MRI配准中的应用8.6 基于非均匀多层自由可变形模型的非刚体配准通用框架结构全英文教材： Rafel C. Gonzalez, Richard E. Woods, Digital Image Processing, Second Edition, Publishing House of Electronics Industry, Beijing, 2002主要参考书：1. 丁明跃，蔡超，医学图像处理，高等教育出版社，2009Syllabus of Graduate Courses in English in College of Li
8、fe Science and TechnologyCourse Name: Medical image analysis Course Code: 170.594Course Type: For Doctoral Students For Master StudentsAssessment: Examination in English Teaching Language: EnglishFor Majors: Bio-medical engineeringLevel： Master DoctoralSemester: Fall Credit Hours: 32 Credits: 2 Prer
9、equisite Courses: medical image processingTeachers Title Major Age Research InterestsDing Mingyue（chief instructor）Professor Bio-medical engineering51 Medical image processing，ultrasound imagingZhang Xumin Associate ProfessorBio-medical engineering37 Medical image processing，ultrasound imagingHou We
10、nguang Lectuer Bio-medical engineering38 Medical image processing，ultrasound imagingOversea Experience and Academic Expertise of the Course Principal Instructor:Ding Mingyue, Professor and Chair of Department of Bio-medical Engineering, Professor and Director of Computer Vision Laboratory, Institute
11、 for Pattern Recognition and Artificial Intelligence, Graduate Director of Institute for Pattern Recognition and Artificial Intelligence, Sept.91-Jul. 93: Research Fellow, Institute for Robotics and Computer Control Technical University of Braunschweig, Braunschweig, Germany. Dec.96-Feb. 97, Jun.99-
12、Aug. 99: Visiting Professor, Technical University of Braunschweig, Braunschweig, Germany. Apr. 2000- Jul. 2000: Research Fellow, Department of Manufacturing Engineering and Engineering Management, City Univ. of Hong Kong, Hong Kong. Nov. 1, 2000-March 2005: Research Associate, Imaging Research Labor
13、atories, Robarts Research Institute, London, Ontario, Canada. During this time, he has held 10 class hours academic seminar for the Robarts Research Institute, in addition to lecturing a 26 class hours of “Modern Medical Image Analysis” course (course code: BME519) in Fall 2004 for the University of
14、 Western Ontario in Canada. Ever since 1985, he has been actively involved with image analysis and processing as well as teaching.Since his return to China in 2005, he has further specialized into medical image processing research and teaching, and is the co-applicant of national natural science fou
15、ndation , National 973, High-tech 863, National Support Program, and other thirty foundations. He has already edited eight Chinese and English publictaions. Publishing in national and international conferences over one hundreds times, of these , 15 has been indexed by SCI, over 100 by EI and over 50
16、 by ISTP. He has also received one European patent, six Chinese patents, three pending American patents. He has received numerous distincitions such as form German DAAD foundation, Canadian Young Investigator Award totaling 23 distinctions. He is also an IEEE Senior Member, SPIE member, Chinese Elec
17、tronic Association, Chinese Bioengineering Senior Memebr, Editor of Electronic Journal, Editor of Aerospace Journal, Chinese Medical Instrument Technology Innovation Alliance Expert Committee Member, Hubei Provincial Biomedical Engineering Medical Intrument Committee Member Course Objectives: Introd
18、uce most often used image segmentation and registration method；Provide an opportunities for students to practice the oral and scientific English, know the concepts and terminologies；Through the full English procedure in lecturing, questions and answers, examinations, improve the reading and usage ab
19、ility of English.Course Outline:Chapter 1 Introduction（2 hours）1.1 What is medical image analysis？1.2 Difference between image processing and image analysis1.3 Most important tasks in medical image analysis: segmentation and registration1.4 Examination compositionChapter 2 Image segmentation based o
20、n neural network （4 hours ）2.1 What is artificial neural network？2.2 Basic idea of neural network based image segmentation 2.3 Image segmentation based on Self-organized neural network2.4 Self-organized neural network image segmentation using texture feture Chapter 3 Deformable models（ 4 hours）3.1 W
21、ho invent the deformable model?3.2 What is deformable models?3.3 Representation of contour3.4 Internal and external force3.5 Deformation of contour3.6 可变模型方法的发展Chapter 4 3D ultrasound image segmentation（4 hours ）4.1 3D ultrasound imaging4.2 Prostate brachytherapy4.3 3D ultrasound image needle segmen
22、tation4.4 3D ultrasound image prostate segmentation4.5 Radio-active seed segmentation in 3D ultrasound imagesChapter 5 Basic concepts of image registration（4 hours）5.1 What is image registration？5.2 Why need to register images？5.3 Basic procedure of image registration5.4 Geometry transformation in i
23、mage registration5.4 Classification of image registration methodsChapter 6 Image registration based on coarse-fine stage （4 hours ）6.1 Two stage registration structure6.2 Image coarse registration based on features6.3 Fine registration based on area6.4 Image registration method based on Harris corne
24、r detectionChapter 7 Image registration using mutual information （6 hours ）7.1 Information measurement7.2 Definition of entropy7.3 What is mutual information？7.4 Properties of mutual information7.5 Structure of image registration based on mutual information7.6 MRI and PET image registration based on
25、 mutual informationChapter 8 Non-rigid image registration (4 hours)8.1 Rigid and non-rigid object8.2 Application of non-rigid transformation8.3 Global and local transformation8.4 Similarity measurement8.5 Non-rigid image registration using free-form8.6 General framework of non-rigid registration based on non-uniform multi-layer free form models Textbook: Rafel C. Gonzalez, Richard E. Woods, Digital Image Processing, Second Edition, Publishing House of Electronics Industry, Beijing, 2002References:1. Ding Mingyue, Cai Chao, Medical image processing, High Education Publisher，2009
