醫學影像分割是一件頗為複雜的工作，必須結合影像處理、電腦圖學及解剖學等多方面的知識才能加以完成。在本論文中，我們將提出如何利用影像處理的技術，配合專家系統中的黑板架構，將使用者所感興趣區域的輪廓線產生出來，並藉由整合本實驗室另一套「以輪廓線為主的三維物件重建系統」，提供使用者在系統做影像辨識的過程中與系統互動的介面及三維物件結果的展示。

在影像處理方面，我們將介紹針對不同的器官所採用的影像處理方法，並建議如何利用「否定性的形狀限制」(negative shape constraints)的概念來幫助我們找出正確的區域，除此之外，我們還提出如何對影像分割之後所產生的輪廓線作評估的方法，並利用顏色的不同來表示結果的好壞及讓使用者作最後輪廓線正確性的確認。

在黑板架構方面，我們將介紹如何將影像及分割結果做集中且一致性的管理，並讓外界能透過相關的存取介面對存放在黑板資料結構內的資料作存取，另外我們還討論如何將影像分割的方法以知識源的形式呈現，並利用統一的介面和控制機制互動以完成影像辨識的工作。

The segmentation of a medical image is an integrated task. We need to integrate the knowledge of image processing, computer vision and anatomy to complete the task. This thesis describes how to use the techniques of image processing with a blackboard architecture to generate the contours of the regions of interest. We also integrate another system, the 3D Builder, to provide the interface for the users so that they can communicate with our system interactively and view the results of the 3-dimensional reconstruction during the process of recognition.

As far as image processing is concerned, we will describe the segmentation methods for different organs and how to use the idea of “negative shape constraints” to help us find out the correct regions. Furthermore, we also describe the methods to evaluate the contours generated by the segmentation knowledge sources and use different colors to show the user about the results of the segmentation.

About the blackboard architecture, we describe how to manage the image data and the segmentation results and provide the interfaces to access them. We also implement the segmentation methods in the from of knowledge sources and provide the interface to enable users to confirm the results of the segmentation.

[Li 94] Hongyi Li, Deklerck, R. and Cornelis, J. ,” Integration of multiple knowledge sources in a system for brain CT-scan interpretation based on the blackboard model”, Artificial Intelligence for Applications, 1994., Proceedings of the Tenth Conference on , 1-4 March 1994, pp. 336-343.
[Masa 95] Masaharu Kobashi and Linda G. Shapiro, “Knowledge-Based Organ Identification From CT Images”, Pattern Recognition, Vol. 28, No. 4, pp. 475-491,1995.
[Brow 97] Matthew S. Brown, Michael F. McNitt-Gray, Member, IEEE, Nicholas J. Mankovich, Jonathan G. Goldin, John Hiller, Laurence S. Wilson, and Denise R. Aberle, “Method for Segmentation Chest CT Image Data Using an Anatomical Model：Preliminary Results”, IEEE Transactions On Medical Imaging, Vol. 16, No. 6, pp. 828-839, December 1997.
[Crev 97] Daniel Crevier and Richard Lepage, “”Knowledge-Based Image Understanding Systems：A Survey”, Computer Vision and Image Understanding, Vol. 67, No. 2 August, pp. 161-185, 1997.
[Kwan 00] F. Y. Kwan, K. C. Cheung, and I. Gibson, “Automatic Extraction of Bone Boundaries from CT Scans Using an Intelligence-based Approach”, Pattern Recognition, Vol. 3, pp. 437-440, Sept. 2000.
[Yoo 00] Seung-Wha Yoo, Jun-Sik Cho, Seung-Mu Noh, Kyung-Suk Shin and Jong-Won Park, “Organ Segmentation by Comparing of Gray Value Portion on Abdominal CT Image”, Proceedings of ICSP2000,pp. 1201-1208.
[Yang 01] Shi-Ming Yang, “An Image Segmentation System Based on Blackboard Architecture”, Master thesis, National Cheng Kung University, Tainan, Taiwan, 2001.
[Chen 03] Chun-Hao Chen, “A Contour-Based 3D Reconstruction System”, Master thesis, National Cheng Kung University, Tainan, Taiwan, 2003.
[鄭慶明 01] 鄭慶明， “實用醫學影像診斷學”，國立編譯館，11月，2001。