透過您的圖書館登入
IP:18.219.224.103

摘要


本研究利用各種不同的影像分析技術,針對多切面電腦斷層獲取的胸腔影像進行處理,精確的描繪出肺部腫瘤區域,希望能夠及早提供資訊,以便進行臨床診斷與治療。研究過程共使用五種複雜程度不等的演算法則[Roberts演算法,Sobel演算法,Laplacian演算法,Marr-Hildreth演算法,Canny演算法]進行肺部腫瘤區域描繪,其中邊界完整性最好的演算法則是Man-Hildreth演算法;可以繪製出的單一圖元邊界和無缺口之封閉區域。而其中以Canny演算法結果與真實肺腫瘤之吻合度最高,也能精確的描匯出肺部腫瘤區域和鄰近胸壁、脊椎、或是縱隔腔等正常組織的界線,是研究時可靠的檢測模式。

關鍵字

影像分割 邊緣檢測 肺腫瘤

並列摘要


Helical computed tomography (CT) is widely used to evaluate numerous lung diseases. Multi-slice CT (MSCT) scanner has further expanded the role of CT in the diagnostic assessment of patients. The large amount of MSCT scan data has promoted many researchers to provide useful and reliable information to assist radiologists in the evaluation of these images. In this paper, five different edge detection algorithms (namely Roberts algorithm, Sobel algorithm, Laplacian algorithm, Marr-Hildreth algorithm and Canny algorithm) have been applied as preprocessing for lung tumor outlining detection. Among all these algorithms, Marr-Hildreth algorithm can provide a complete contour of lung tumor outline and the Canny algorithm eventually achieve the best outline contour.

並列關鍵字

image segmentation edge detection lung tumor

延伸閱讀