This course has been designed as a topic-driven learning course. The focus of this semester is on 2D and 3D segmentation techniques for various medical applications. Segmentation is an essential process for most quantitative medical image analyses, e.g., tumor characterization, brain function analysis, vessel analysis, etc. The classes will be divided into two parts. In the first part, the instructor will present background knowledge for such widely-used techniques as deformable models, level set methods, watershed transformation based approaches, wavelet transformation based approaches, morphological operators, and so on. In the second part, each student will be required to present a paper on image segmentation. Furthermore, the students will be teamed up into two or three groups, each of which needs to solve a practical medical problem by implementing the state-of-the-art segmentation techniques.

持續改善：上學期主要以概念式的方法與方向來進行課程內容教學，本學期透過不同方式針對主題式的方向與方法進行課程的調整與異動符合實際需求，可配合全球此領域研究方向的多變，更容易配合不同領域學生進行學習。針對主題技術(EX：Fuzzy c mean、watershed、cut...etc)可直接了解不同方法間常運用的概念，並可快速針對不同主題在新的研究方法與技術上延伸教學，透過不同主題技術的學習如何直接應用在不同醫學影像技術上面，配合不同醫學影像技術達到更有效率的學習與應用。