臺灣博碩士論文加值系統

在工業化國家由於人民生活壓力增加，許多疾病都逐漸威脅人民健康，尤其在心血管疾病方面。而國內外各種影像技術在心血管診治工作中均佔有舉足輕重的地位，而核磁共振（Magnetic Resonance Image）是其中一個重要組成部分。但由於靜態血流信號可使心肌與血液的界線模糊不清，所以醫師在心室內外膜之邊緣檢測程序上，依然以手動方式來圈選其邊界，不僅非常耗時且無法持續其準確性及醫療品質。本研究以心臟核磁共振影像為案例，強化左心室之資訊，進而自動圈選內外膜之邊界，提供醫師一有效之診斷工具。
本研究分為兩階段，第一階段為應用小波變換法來強化左心室之影像資訊。小波變換法發展歷史雖不久，但在強化影像技術上卻有不錯之功效。第二階段為結合 Fleagle et. al. [7] 之心室邊界檢測法，進而評估本研究與其演算法在內外心膜邊界檢測上之績效。
本研究可達成下列具體成果：(一) 提供醫師一自動檢測內外心膜之演算法，以減輕醫師之負擔，進而維持醫療品質。(二) 提供有關影像強化技術於心臟核磁共振影像上，以供後續研究之參考。

MRI system is noninvasive and provides the clear images to diagnosis. In cardiovascular system, however, MR images require manual trace method to identify the endocardial border and the epicardial border in left ventricular. Because dynamic organs generate a huge number of images, it takes long time to identify them by using the manual trace method. To provide satisfactory clinical performance, an automatic endocardial and the epicardial border detection algorithm is required.
In this research, we provide an algorithm of wavelet-based images enhancement. One hundred and sixty images from ten volunteers and divide into three groups：(1)：borders are manual tracing as a compare group, (2)：the automatic border detection algorithm is directly without images enhancement .
(3)：the automatic border detection algorithm was applied after the images are enhanced by WT-based method. Finally we use the Hausdorff Distance to measure the performance of the images with or without image enhancement.
Experimental results show that the endocardial profiles and the epicardial profiles can be effectively enhanced by the wavelet-based technique.

第一章 緒論.....................................1
1.1 研究背景及動機.........................1
1.2 研究目的...............................2
1.3 研究範圍...............................2
第二章 文獻探討.................................4
2.1 參考文獻...............................4
2.2 相關文獻探討...........................5
2.3 主要文獻探討...........................6
2.3.1 階梯平滑化........................6
2.3.2 小波變換..........................6
2.3.3 動態規劃..........................9
第三章 研究架構與方法..........................11
3.1 研究架構流程..........................11
3.2 研究方法..............................14
3.2.1 階梯平滑化.......................14
3.2.2 小波變換.........................15
3.2.3 邊界檢測.........................24
3.2.4 摺積運算.........................24
3.2.5 動態規劃.........................25
3.2.6 Hausdorff Distance...............27
第四章 實驗結果與分析..........................29
4.1 實驗結果..............................29
4.2 統計分析..............................35
4.2.1 Hausdorff Distance 樣本檢定......35
4.2.2 變異數檢定.......................37
第五章 結論與未來發展..........................41
5.1 結論..................................41
5.2 未來發展..............................41
參考文獻 42
附件A 實驗組之Hausdorff Distance 數據.........44
附件B 受測者之實驗輸出影像...................54
附件C 已接受之論文
a. 傅家啟、連漢仲、劉邦彥，1999年3月，自動化剔除
人為動作干擾功能之電腦輔助胃電圖診斷系統開發，大葉學報
b. 傅家啟、蔡志文、劉邦彥，1999年5月，核磁共振影
像左心室短軸面之邊界強化與檢測，中西醫學應用研討會論文集

1. Atkins, M. S., Mackiewich, B.T., Fully Automatic Segmentation of the Brain in MRI,IEEE Transactions
on Medical Imaging, Vol. 17,No. 1, 1998 , pp. 98-107.
2. Chan, H. P., et. al., Improvement in Radiologists'' Detection of Clustered Microcalcifications on Mammograms — The Potential of Computer-Aided Diagnosis. Invest. Radiology, No. 10, Oct., 1990, pp. 1102 — 1110.
3. Fleagle, S.R., Thedens, D.R., Stanford, W., Pettigrew, R.L., Reichek, N.,Skorton, D. J, Multicenter Trial of Automated Border Detection in Cardiac MR Imaging, JMRI , Vol. 3, No2.March / April 1993.
4. Giger, M. L, et. al., An Intelligent Workstation for Computer-AidedDiagnosis, Radio. Graphics, Vol. 13. 1993, pp. 647 - 656.
5. Gordon, R., Rangayyan, R. M., Feature enhancement of film mammograms using fixed and adaptive neighborhoods, Applied Optics, Vol. 23, No. 4, 1984, pp. 560 - 564.
6. Gramatikov, B. and Georgiev, I. Wavelets and Alternative to Short-Time Fourier Transform in Signal-Averaged Electrocardiography, Med. & Bio. Engr. & Computig, Vol. 33, No. 3, 1995, pp. 482 — 487.
7. Laine, A. F. et. al., Mammographic Feature Enhancement by Multiscale Analysis, IEEE Transactions on Medical Imaging, Vol. 13, No. 4, December 1994, pp. 725 — 740
8. Laine Andrew F., Sergio Schuler, Jian Fan, and Walter Huda Mammographic Feature Enhancement by Multiscale Analysis, IEEE Transactions on Medical Imaging, Vol. 13, No. 4, 1994, pp. 725-740.
9. Lin, Zhiyue and Chen, J. Z. Chen, Comparison of Three Running Spectral Analysis Methods, Electrogastrography Principles and Applications, 1994, pp. 75-99.
10. Mallat, S.,A wavelet tour of signal processing, Academic Press,1998.
11. Nishikawa, R. M., et. al., Computer-Aided Detection of Clustered Microcalcifications - An Improved Method for Grouping Detected Signals, Med. Physics, 20 (6), 1993, pp. 1661 - 1666.
12. Nishikawa, R. M., et. al., Computer-Aided Detection of Clustered Microcalicifications on Digital Mammograms, Med. & Bio. Engr. & Computig, March, 1995, pp. 174 - 178.
13. Qian Wei, Maria Kallergi, Laurence P. Clarke, Huai-Dong Li, Priya Venugopal, Dnsheng Song, and Robert A. Clark, Tree structured wavelet transform segmentation of microcalcifications in digital mammography, Medical Physics, Vol. 22, No. 8, 1995, pp. 1247-1254
14. Vikram Chalana and Yongmin Kim, A Methodology for Evaluation of Boundary Detection Algorithms on Medical Images, IEEE Transaction on Medical Images ,Vol. 16 ,No 5 Oct 1997, pp642-652