X光影像之自動化肺部切割與病變細胞辨識之研究

Research of Automatic Lung Segmentation and Morbid Cell Recognition in X-ray Images

陳柏揚(Po-Yang Chen)

指導教授 ： 洪國龍

朝陽科技大學/資訊學院/資訊管理系/碩士(2009年)

摘要

肺在人體中佔有非常重要的地位，當肺部發生病變時，醫生必須使用多種不同的顯影技術來檢查肺部的病變區域，可得到更詳細、更精確的結果。最常使用的攝影器材有電腦斷層造影（Computed Tomography, CT）、核磁共振造影（Magnetic Resonance Imaging, MRI）、X光造影（X-rays）等等。本研究採用X光造影所得到之胸腔影像進行實驗，影像處理技術在醫學影像領域中亦能加以應用，本篇方法在切割描繪肺域邊緣上採用Canny邊緣偵測、連通區域標記與Convex Hull邊緣描繪等方法進行自動化肺葉邊緣描繪。
肺部區域描繪完成之後，使用影像分割與粗糙度特徵分析，將肺葉內各區域進行分析與辨識，為了提昇病變細胞候選者的邊緣，先進行Sobel加強邊緣處理，爾後利用粗糙度值進行排序篩選，尋找出病變細胞候選者並框選出來，提供醫療人員進一步的參考。

關鍵字

病變細胞 ； 肺部切割 ； 灰階門檻值 ； 粗糙度

並列摘要

The lung holds the very important status in the human body, when the lungs have the pathological change, doctor must use many kinds of different development technologies to inspect lungs'' pathological change region, may obtain is more detailed, a more precise result. Most often uses the photographic equipment has the computer fault radiography (Computed Tomography, CT), the nuclear magnetic resonance radiography (Magnetic Resonance Imaging, MRI), X light radiography (X-rays) and so on. This research uses X light radiography to obtain the chest cavity phantom to carry on the experiment, the image process technology can also perform in the medicine phantom domain to apply, this method uses the Canny edge detection, the connected domain mark and Convex in the cutting description lung territory edge methods and so on Hull edge description carries on the automated lobe of the lung edge description.
After the lungs region segmentation, the use phantom division and roughness characteristic analysis, various regions carries on the lobe of the lung in the analysis and the identification, to promote the morbid cell candidate''s edge, carries on Sobel to strengthen the edge processing first, in future will carry on sorting using roughness value to screen, seeks for the morbid cell candidates and the frame elects, will provide douctors further reference.

並列關鍵字

Gray Scale Threshold Value ； Roughness ； Morbid Cell ； Lungs Segmentation

參考文獻

[4] S. G. Armato, III, M. L. Giger, and H. MacMahon, “Automated lung segmentation in digitized posteroanterior chest radiographs”, Academic Radiology, vol. 5, no. 4, pp. 245-255, 1998.

[5] R. Boscolo, M. S. Brownand, and M. F. McNitt-Gray, “Medical Image Segmentation with Knowledge-guided Robust Active Contours”, Radiographics, vol. 22, pp. 437-448, 2002.

[6] M. S. Brown, L.S. Wilson, B. D. Doust, R. W. Gill, C. Sun, “Knowledge-based method for segmentation and analysisof lung boundaeries in chest X-ray images ”, Computerized Medical Imaging and Graphics, 1998 pp.463-477.

[8] M. Brown, R. Gill, H. Talhami, L. Wilson, and B. Doust, “Model-based assessment of lung structures: Inferencing and control system”, Proceedings of SPIE Medical Imaging, San Diego, CA, vol. 2433, pp. 167-178, 1995.

[9] L. D. Cohen, “On active contour models and balloons”, Computer Vision, Graphics, and Image Processing: Image Understanding, vol. 53, no. 2, pp.211-218, 1991.