智慧型初期缺血性腦中風偵測系統

在本論文中，我們提出了一套智慧型初期缺血性腦中風偵測系統，它能夠輔助醫師診斷，使原本不易看出的腦中風區域更有視覺上的感知增強。本論文分成三個主要的部份。首先，將輸入的電腦斷層掃描(Computed Tomography, CT)影像作前處理使之利於後續的步驟使用，並利用數學形態學提取大腦組織的部分。接著，是使用自動化的區域生長算法分割腦組織區域。這是為了在隨後的第三階段中可以明確區分白質和灰質為八個區域。第三階段，本系統使用所提出的簡易區域亮度比較算法，來找到一致的大腦區域位置的方法。目的在找到腦中風區域位於右半腦或左半腦後，將進行最後的修正和白色標記的面積就是本系統執行完預測的腦中風區域。最後，為了驗證本系統所提出的方法，本文將程序執行完的結果經過實際的測試，先由本校附設醫院的專業醫師的幫助將有中風的影像出取出來，再透過兩位沒有經驗的放射科醫生進行檢測，比較是否本系統能夠真正的幫助到醫師來提升腦中風診斷的正確率，實驗數據證明成功率由原本的最低33％經由本系統輔助後，醫師的結果有超過一半進行了更動，並且經由更動後最終成功率可達到64%以上，系統偵測的靈敏度經測試後也達到了85.55%，且在Object-level Consistency Error(OCE)的評估上其成功率達到了0.8678。

關鍵字

缺血性腦中風；電腦斷層掃描；數學形態學；區域生長； Object-level Consistency Error

並列摘要

In this thesis, we propose an intelligent early ischemic stroke detection system. It can help doctor to diagnosis. The system will be divided into three parts. First, the input Computed Tomography (CT) image will be performed preprocessing. The preprocessing step includes contrast enhancement using cubic curve, and extraction brain tissue using morphology technology in the image processing. The second, the brain tissue area will be segmented using Unsupervised Features region growing algorithm (UFRGA). Its goal is to distinguish the white matter and gray matter in the brain area. Finally, we use a simple position coincidence method to coincide the intensity according to the intensity of areas obtained by our proposed method. Hence, the brain stroke area will be found out. In the experiment result, we invite the two radiologists to help us to test our proposed system. And, the three statistic indices and an empirical evaluation index are used to evaluate our proposed system. From the result, we know that our proposed system can aided radiologist to increase themselves success rate to 64% up and the sensitivity of the system is 85.55%. The empirical evaluation index (Object-level Consistency Error, OCE) is 0.8678.

並列關鍵字

Ischemic Stroke ； Computed Tomography ； Mathematical Morphology ； Region Growing ； Object-level Consistency Error

參考文獻

[9] Jing-Jia Lin, Hsin-Ling Cheng and Tai-Yue Wang,“ Predicting Recurrent Stroke via ANN Model,” The Journal of Taiwan Assocation for Medical Informatics, Taiwan, 2010.

[20] Chin-Teng Lin and Chiun-Li Chin,“ Using Fuzzy Inference and Cubic Curve to Detect and Compensate Backlight Image,” International Journal of Fuzzy Systems, vol. 8, pp. 2-13, 2006.

[1] Yoshida H. and Nappi J.,“ Three-dimensional computer-aided diagnosis scheme for detection of colonic polyps,” IEEE Transactions on Medical Imaging, pp. 1261-1274, 2001.

[2] Hiroyuki Yoshida, Janne Näppi and Peter MacEneaney et al.,“ Computer-aided Diagnosis Scheme for Detection of Polyps at CT Colonography,” RadioGraphics, pp. 963-979, 2002.

[3] Kunio Doi,“ Computer-Aided Diagnosis in Medical Imaging: Historical Review, Current Status and Future Potential,” Comput Med Imaging Graph, pp. 198-211, 2008.

國際替代計量

智慧型初期缺血性腦中風偵測系統

未授權

主題瀏覽