應用藥物動力學的乳房DCE-MRI電腦輔助診斷

Computer-aided Diagnosis of Breast DCE-MRI Using Pharmacokinetic Model

張倍綱(PEI-KANG CHANG)

指導教授 ： 張瑞峰

國立臺灣大學/電機資訊學院/資訊工程學研究所/碩士(2009年)

https://doi.org/10.6342/NTU.2009.10182

摘要

乳癌是女性第二大容易罹患的癌症，並且是全球女性的主要死因之一。但是乳癌也是一種可以早期發現、早期治療的疾病，並且因此有相當大的機會可以治癒。近來，電腦輔助診斷持續不斷地發展；對於乳癌的檢測上，它不僅能提供腫瘤的資訊，甚至能近一步地偵測出腫瘤的位置、大小，以及分辨出腫瘤的良惡性，於是可以有效地減少像是腫瘤切片等等的侵入性檢查方式。在本篇論文內，動態對比增強的核磁共振影像被使用作為診斷的工具，其可記錄打入顯影劑後，腫瘤訊號隨時間的變化。我們將提出一個新的利用動力學彩色圖及曲線下面積彩色圖的切割方法以找出腫瘤，同時再利用模糊C-means及平均二種方法找出一條動力學曲線來代表一個腫瘤，接著將使用藥物動力學的數學模型來分析此動力學曲線，其所得的參數是腫瘤良惡性判斷的根據，提供作為實驗統計分析上的資料。在實驗中，除了藥物動力學的模型，還使用了指數型態的數學模型和傳統常用的動力學特徵，並將這三類特徵分別測試其腫瘤良惡性分類的效果。在我們的實驗資料中有總計124個病理檢驗過的腫瘤，其中包含78個惡性的、46個良性的病例。根據實驗的結果，發現由藥物動力學模型所得的特徵最能區別出腫瘤的良惡性，能達到準確性79.84%、敏感性80.77%、專一性78.26%及Az值0.8604。

關鍵字

電腦輔助診斷 ； 藥物動力學 ； 動態對比增強 ； 乳房核磁共振影像 ； 彩色圖 ； 動力學 ； 曲線下面積

並列摘要

The breast cancer is the second most common cancer and the major cause of death for women. However, it is also a type of cancer that could be early detected and has an excellent curability in early stage. Recently, the computer-aided diagnosis (CAD) system develops persistently. For breast cancer, the CAD system provides the information about the tumor for radiologists and further detects the lesion and differentiates the benignancy from malignancy of the tumor. Thus, the invasive examination, like the biopsy, might be reduced. In this study, the dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is used to record the change of the signal intensity of the tumor over time after injecting the contrast agent by vien. In this paper, a novel segmentation method using the kinetic color map and the area-under-the-curve color map is proposed to find the tumor. The fuzzy c-means clustering and the kinetic curve average are used to identify a kinetic curve of the tumor for analysis. The pharmacokinetic model is used to fit the kinetic curve of the tumor. The parameters of the fitted pharmacokinetic model are used as the diagnosis features for statistic analysis and they will be compared with the features of the exponential model and conventional kinetic curve characteristics. A total 124 lesions with 78 malignant and 46 benignant are included in our study. The result of our experiment shows that the feature set of pharmacokinetic model has better and stable performance. Its accuracy, sensitivity, specificity, and Az value are 79.84%, 80.77%, 78.26%, and 0.8604, respectively.

並列關鍵字

CAD ； DCE-MRI ； Breast ； Color Map ； AUC ； Kinetic ； Pharmacokinetic

參考文獻

[1] D. Saslow, C. Boetes, W. Burke, S. Harms, M. O. Leach, C. D. Lehman, E. Morris, E. Pisano, M. Schnall, S. Sener, R. A. Smith, E. Warner, M. Yaffe, K. S. Andrews, and C. A. Russell, "American Cancer Society guidelines for breast screening with MRI as an adjunct to mammography," CA Cancer J Clin, vol. 57, pp. 75-89, Mar-Apr 2007.

[3] W. A. Berg, J. D. Blume, J. B. Cormack, E. B. Mendelson, D. Lehrer, M. Bohm-Velez, E. D. Pisano, R. A. Jong, W. P. Evans, M. J. Morton, M. C. Mahoney, L. H. Larsen, R. G. Barr, D. M. Farria, H. S. Marques, and K. Boparai, "Combined screening with ultrasound and mammography vs mammography alone in women at elevated risk of breast cancer," JAMA, vol. 299, pp. 2151-63, May 14 2008.

[4] W. A. Berg, L. Gutierrez, M. S. NessAiver, W. B. Carter, M. Bhargavan, R. S. Lewis, and O. B. Ioffe, "Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer," Radiology, vol. 233, pp. 830-49, Dec 2004.

[5] D. A. Bluemke, C. A. Gatsonis, M. H. Chen, G. A. DeAngelis, N. DeBruhl, S. Harms, S. H. Heywang-Kobrunner, N. Hylton, C. K. Kuhl, C. Lehman, E. D. Pisano, P. Causer, S. J. Schnitt, S. F. Smazal, C. B. Stelling, P. T. Weatherall, and M. D. Schnall, "Magnetic resonance imaging of the breast prior to biopsy," JAMA, vol. 292, pp. 2735-42, Dec 8 2004.

[6] U. Fischer, L. Kopka, and E. Grabbe, "Breast carcinoma: Effect of preoperative contrast-enhanced MR imaging on the therapeutic approach," Radiology, vol. 213, pp. 881-8, Dec 1999.