基於肺腫瘤電腦輔助系統診斷架構下之肺區擷取實作

Implementation of Lung Extraction Based on the Structure of Lung Nodule Computer Aided Diagnosis System

蔡祐慎(Yu-Shen Tsai)

指導教授 ： 盧以詮

元智大學/資訊學院/資訊管理學系/碩士(2011年)

摘要

由於醫學影像分割技術的發展與進步，CAD的應用愈來愈普遍，其主要之目的為協助醫師識別出具有疑似病因意義的影像，以利後續之診斷與追蹤。以肺腫瘤檢測為例，其功能是協助專科醫師在數位化電腦斷層胸腔影像中，以識別、鑑定及評估肺部病灶與結節，使得肺部疾病可以提早發現，及早治療。然而事實上，CAD的建置相當費時，因為其背後所包含的理論相當廣泛，而實作起來更是格外困難。目前CAD的建置理論一般都是以基於規則的方法(Rule-based)和類神經網路(Neural Network)兩者為主。本文從蒐集資料開始，嘗試建構一個完善的CAD去對腫瘤做有效的判斷，但礙於開發時間問題，本文僅實作CAD的第一步，肺區擷取(Extraction of lung region)。肺區擷取的好壞深深地影響診斷的結果，因此，若能在肺區擷取上取得良好的效果，對於實作CAD的後面步驟而言，將會更有效率。

關鍵字

肺區擷取 ； 肺腫瘤電腦輔助判斷系統 ； 類神經網路

並列摘要

The primary function of the CAD system is to help doctors identify those images with suspicious pathological lung regions, make precise diagnoses, and successfully cure lung cancer patients. Therefore, this thesis has collected a great number of correlated theories and research papers, aims to construct a complete structure of the CAD system, which is usually supported by a Rule-based system or Neural Network architectures, for doctors and radiologists to make effective diagnoses of lung cancer patients. However, limited by time and money, this thesis only operates the first step of the CAD system, which is the lung region extraction. Although this thesis only focuses on the lung region extraction, and thus cannot execute every step of the CAD system, it can still serve as a useful and valuable example for those who are interested in the implementation of the following steps of the CAD system. In fact, the results of lung region extraction can deeply influence later diagnoses and evaluations. Consequently, if lung region extraction can be successfully accomplished, it will definitely help the following steps of the CAD system implementation work more functionally.

並列關鍵字

lung region extraction ； the CAD system for lung nodule detection ； Neural Network

參考文獻

Armato, S. G., and Sensakovic, W. F., “Automated lung segmentation for thoracic CT,” Academic Radiology, vol.11, pp. 1011–1021, September 2009.

Armato, S. G., Giger, M. L., and MacMahon, H., “Automated detection of lung nodules in CT scans: preliminary results,” Medical Physics vol.28, pp.1552–1561, 2001

Awai, K., Murao, K., Ozawa, A., Komi, M., Hayakawa, H., Hori, S., and Nishimura, Y., “Pulmonary nodules at chest CT: effect of computer-aided diagnosis on radiologists’ detection performance,” Radiology, vol.230, pp.347–352, February 2004.

Bankman, I. N.,“ Segmentation algorithms for detecting micro calcifications,” IEEE Transactions on Medical Imaging, vol.11, pp.141-149, February 1996

Brown, M. S., McNitt-Cray, M. F., Golldin, J. G., Suh, R. D., Sayre, J. W. and Aberle, D. R., “Patient-specific models for lung nodule detection and surveillance in CT images,” IEEE Transactions on Medical Image, vol.20, pp.1242–1250, December 2001.