肺動脈栓塞(Pulmonary Embolism, PE)是一種肺動脈或其細支血管被栓子堵住的現象。通常是由靜脈中的血栓隨著血液的流動而到了肺動脈中，並且阻塞了肺動脈。傳統的偵測都必須藉重醫師們的專業知識，肺栓塞嚴重的時候會導致病人的死亡，因此在診斷肺栓塞的部份是相當重要的一個環節。本論文使用CT血管攝影(Computed Tomography Angiography, CTA)影像來做偵測，首先我們先使用ACM的方法將前處理影像中肺部的位置擷取出來，並且針對肺部的血管影像進行搜尋與追蹤的動作，同時利用栓塞所俱有的特徵去偵測肺動脈或其細支血管中是否有栓塞。最後，我們使用的樣板比對（Template matching）的方法將醫師所圈選出來的結果和我們所提出方法的結果進行比對，以此法來計算我們方法的準確率。實驗結果顯示系統的偵測成功率可達81%，同時系統能提供肺栓塞的位置資訊給醫師，做為輔助醫師診斷時的依據，同時也減輕醫師的負擔。

Pulmonary embolism (PE) is a blockage of the pulmonary artery or one of its branches, usually occurring when a deep vein thrombus (blood clot from a vein) becomes dislodged from its site of formation and travels, or embolisms, to the arterial blood supply of one of the lungs. This process is termed embolism. Most of the traditional detection of PE needs to rely on the professional judgments of physicians. With advances in CT technology, faster scanning and greatly enhanced the quality of image improve the diagnostic accuracy. Serious PE will lead to dead. However, it is very important for diagnosis PE. First, we use ACM method to find the lung area of the preprocessed image. And then we perform vessel searching and tracking to detect the pulmonary embolism in lung. Finally, we use the template matching method to do comparison between compare the result of system and doctor’s and obtain the precision rate. The system shows the result to help doctors checking the position of the embolism and release their work loads.

中文摘要 ................................................................................................................................... I
英文摘要 .................................................................................................................................. II
致謝 ........................................................................................................................................ III
Table of Contents ...................................................................................................................... V
Table of Figures ....................................................................................................................... VI
Table of Tables ....................................................................................................................... VII
Chapter 1 Introduction ....................................................................................................... 1
1.1 Foreword .................................................................................................................. 1
1.2 Motivation researching and background .................................................................. 2
1.3 Thesis arrangement .................................................................................................. 8
Chapter 2 Methodology ..................................................................................................... 10
2.1 Brief Introduction ................................................................................................... 10
2.2 Pre-processing ........................................................................................................ 13
2.3 Finding Pulmonary Location ................................................................................. 14
2.4 Vessel Searching .................................................................................................... 19
2.4.1 Branch vessel searching ..................................................................................... 22
2.4.2 Main vessel searching ........................................................................................ 25
2.5 Vessel Tracking...................................................................................................... 36
2.6 Precision Computing .............................................................................................. 41
Chapter 3 Experimental Results and Discussion ............................................................ 44
3.1 The introduction of data set ................................................................................... 44
3.2 Define the performance of the system ................................................................... 45
3.3 Experimental results ............................................................................................... 45
Chapter 4 Conclusions and Future works ....................................................................... 52
References .............................................................................................................................. 54

[1] Qanadi SD, Hajjam ME and Mesurolle B, “Pulmonary embolism detection: Prospective evaluation of dual section helical CT versus selective pulmonary arteriography,” Radioloay, Vol. 217, pp. 447-455, 2000. [2] Michiels JJ, Hoogsteden H and Pattynama PM, “Non-invasive diagnosis of pulmonary embolism,” Acta Chir Belg, Vol. 1, pp. 26-34, 2005. [3] Nicholas J. Screaton and Harvey O. Coxson, “Detection of lung perfusion abnormalities using computed tomography in a porcine model of pulmonary embolism,” Journal of Thoracic Imaging, Vol. 18, pp. 14-20, 2003. [4] U. Joseph Schoeph and Nicolaus Holzknecht, “Subsegmental pulmonary emboli: improved detection with thin-collimation multi-detector row spiral CT,” Radiology, pp. 483-490, 2002. [5] Reinhard Groell and Karl H. Peichel, “Computed tomography densitometry of the lung: a method to assess perfusion defects in acute pulmonary embolism,” European Journal of Radiology, Vol. 32, pp. 192-196, 1999. [6] Joachim E. Wildberger and U. Joseph Schoepf, “Approaches to CT perfusion imaging in pulmonary embolism,” Seminar Roentgenol, Vol. 40, pp. 64-73, 2005. [7] Peter Herzog and Joachim E. Wildberger, “CT perfusion imaging of the lung in pulmonary embolism,” Academic Radiology, Vol. 10, pp. 1132-1146, 2003. [8] Wildberger JE, Niethammer MU and Klotz E, “Multi-slice CT for visualization of pulmonary embolism using perfusion weighted color maps,” Fortschr Rontgenstr, Vol. 173, pp. 289-294, 2001. [9] Simon J. Scurrell, “Automatic detection of pulmonary embolism using computation intelligence,” A dissertation of university of the Witwatersrand, October 2006. [10] Schoepf UJ and Costello P, “CT angiography for diagnosis of pulmonary embolism: state of the art,” Radiology, Vol. 230, pp. 329-337, 2004. [11] A. Frigyesi, “An automated method for the detection of pulmonary embolism in V/Q scans,” Medical Image Analysis, Vol. 7, No. 7, pp. 341–349, 2003. [12] R. Damper and I. Middleton, “Segmentation of magnetic resonance images using a combination of neural networks and active contour models,” Medical Engineering & Physics, Vol. 26, No. 1, pp. 71–86, January 2003. [13] Rafael C. Gonzalez and Richard E. Woods, “Digital Image Processing,” Addison-Wesley Publishing Company, 1993.
[14] Margarida Silveira and Jorge Marques, “Automatic segmentation of the
55
lungs using multiple active contours and outlier model,” IEEE Engineering in Medicine and Biology Society Annual International Conference, pp. 3122-3125, 2006. [15] D. Xu, J. Lee, D.S. Raicu, J.D. Furst and D.Channin, “Texture Classification of Normal Tissues in Computed Tomography,” The 2005 Annual Meeting of the Society for Computer Applications in Radiology, Orlando, Florida, June 2-5, 2005. [16] Park W., Hoffman E.A. and Sonka M, ”Segmentation of intrathoracic airway trees: a fuzzy logic approach,” IEEE Transaction on Medical Imaging, vol. 17, No. 8, pp. 489–197, Aug. 1998.
[17] Blechschmidt R.A., Werthschutzky R and Lorcher U, “Automated CT image evaluation of the lung: a morphology-based concept,” IEEE Transactions on Medical Imaging, Vol. 20, No.5, pp. 434-442, 2001. [18] Brown MS, McNitt-Gray MF, Mankovich NJ, Goldin JG, Hiller j, Wilson LS and Aberle DR., “Method for segmentation chest CT image data using an anatomical model: preliminart results,” IEEE Transactions on Medical Imaging, Vol. 15, No. 6, pp. 828-839 , 1997. [19] Dajnowiex, M. and Alirezaie, J., “Computer simulation for segmentation of lung nodules in CT images,” IEEE International Conference on Systems Man and Cybernetics, Vol. 5, pp. 4491-4496, 2004.
[20] Yongbum Lee, Takeshi Hara, Hiroshi Fujita, Shigeki Itoh and Takeo Ishigaki, “Automated detection of pulmonary nodules in helical CT images based on an improved template-matching technique,” IEEE Transactions on Medical Imaging, Vol. 20, No. 7, pp. 595-604 , 2001. [21] Augusto Silva, José S. Silva, Beatriz S. Santos and Carlos Ferreira, “Fast Pulmonary Contour Extraction in X-ray CT images: A Methodology and Quality Assessment,” Medical Imaging 2001: Physiology and Function from Multidimensional Images, Vol. 4321, pp. 216-224, 2001. [22] Hu s., Hoffman E. A., Reinhardt and J. M., “Automatic lung segmentation for accurate quantitation of volumetric X-ray CT images,” IEEE Transactions on Medical Imaging, Vol. 20, No. 6, pp. 490-498 , 2001.
[23] Shojaii, R. Alirezaie and J. Babyn, P., “Automatic lung segmentation in CT images using Watershed Transform,” IEEE International Conference on Image Processing, Vol. 2, pp. 1270-1273, 2005.
[24] U.J. Schoepf, N. Holzknecht, T.K. Helmberger, et al., “Subsegmental pulmonary emboli: improved detection with thin-collimation multi-detector row spiral CT,” Radiology 222 (2), pp. 483-490, 2002.
56
[25] Y. Masutani, H. MacMahon and K. Doi, “Computerized detection of pulmonary embolism in spiral CT angiography based on volumetric image analysis,” IEEE Transactions on Medical Imaging, Vol. 21, Issue 12, pp. 1517-1523, 2002. [26] C. Zhou, et al., “Computerized detection of pulmonary embolism in 3D computed tomographic (CT) images: vessel tracking and segmentation techniques,” In Medical Imaging 2003: Image Processing. Proceedings of the SPIE, Vol. 5032, pp. 1613-1620, May 2003. [27] E.A. Chiang, et al., “Detection of pulmonary embolism: comparion of paddlewheel and coronal CT reformations-initial experience,” Radiology 228 (2), pp. 577-582, 2003. [28] A.P. Kiraly, D.P. Naidich, and C.L. Novak, “Cartwheel projections of segmented pulmonary vasculature for the detection of pulmonary embolism,” Medical Imaging 2005:Image guided procedures, and display. Proceedings of SPIE - Int''l. Society for Optical Engineering, Vol. 5733, Part 1, pp. 69-78, 2005. [29] E. Pichon, C.L. Novak, A.P. Kiraly and D.P. Naidich, “A novel method for pulmonary emboli visualization from high-resolution CT images,” SPIE Medical Imaging, Vol. 5367, pp.161-170, 2004. [30] A.P. Kiraly, C.L. Novak and D.P. Naidich, “A visualization-Based Method for Pulmonary Emboli Identification within High-Resolusion CT Images,” European Congress of Radiology, 2005. [31] Szeliski, R., Tonnesen, D. and Terzopoilos D., “Modeling surfaces of arbitrary topology with dynamic particles,” IEEE Conference on Computer Vision and Pattern Recognition, pp. 82-87, 1993. [32] A.P. Kiraly, “3D Image Analysis and Visualization of Tubular Structures,” PhD thesis, The Pennsylvania State University, 2003. [33] A.P. Kiraly, D. P. Naidich and C. L. Novak, “2D display of a 3D tree for pulmonary embolism detection,” Computer-aided Radiology and Surgery - CARS 2005 - Proceedings of the 19th Int''l. Congress and Exhibition., Vol. 1281, pp. 1132-1136, 2005. [34] Henri Bouma, Jeroen J. Sonnemans, Anna Vilanova and Frans A. Gerritsen, “Automatic Detection of Pulmonary Embolism in CTA Images,” IEEE Transactions on Medical Imaging, Vol. 28, No. 8,August 2009. [35] Michael Kass, Andrew Witkin and Demetri Terzopoulos, “Snakes: Active Contour Models,” International Journal of Computer Vision, pp. 321-331, 1988.