使用拓樸分水嶺切割於全乳房自動超音波的多維度腫瘤偵測

Multi-dimensional Tumor Detection in Automated Whole Breast Ultrasound using Topographic Watershed

洪明仁(Ming-Jen Hung)

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(2013年)

https://doi.org/10.6342/NTU.2013.10297

摘要

全乳房自動超音波是一種熱門的檢測乳房腫瘤的儀器。相較於傳統的手持式超音波，全乳房自動超音波可以達到不依賴操作者的經驗以及用在大規模篩檢的可行性。然而逐張檢查數百片的全乳房自動超音波影像需要花費很多的時間，本論文提出一個基於分水嶺切割的電腦輔助乳房腫瘤偵測系統來加速檢查影像的時間。分水嶺切割會將影像中每一個局部極小灰階值附近相似的組織合併成為一個同質性的區域。利用每一個區域中量化的形態學、灰階值以及材質特徵來計算出此區域被視為腫瘤的可能性，而這三大類的特徵也應用於多維度的偽陽性篩除，來減少非腫瘤區域被誤判成腫瘤的情形發生。實驗中所用到的測試資料包含了68個良性以及65個惡性腫瘤。經由實驗結果，本系統在腫瘤偵測率100% (133/133)，90% (121/133) 以及80% (107/133) 的情況下，平均每一個全乳房自動超音波掃瞄會有9.44，5.42以及3.33個非腫瘤區域被誤判成腫瘤。另外，當系統同時使用三大類的特徵時，品質因數 (FOM) 可以達到0.46，這和使用其他特徵的組合相比時有顯著性差異(p-value<0.05)。總結，本論文所提出的基於分水嶺切割的電腦輔助乳房腫瘤偵測系統可利用多維度的偽陽性篩除有效地偵測在全乳房自動超音波中的腫瘤。

關鍵字

乳癌 ； 電腦輔助乳房腫瘤偵測 ； 全乳房自動超音波 ； 分水嶺切割 ； 多維度的偽陽性篩除

並列摘要

Automated whole breast ultrasound (ABUS) is becoming a popular screening modality for whole breast examination. Compared to conventional hand-held ultrasound, ABUS could achieve operator-independent and is feasible for mass screening. Because reviewing hundreds of slices in an ABUS image volume is time-consuming, a computer-aided detection (CADe) system based on watershed transform was proposed to accelerate the reviewing. The watershed transform was applied to gather similar tissues around local minima to be homogeneous regions. The likelihoods of being a tumor of the regions were estimated using the quantitative morphology, intensity, and texture features in 2-D/3-D false positive reduction (FPR). The collected database comprised 68 benign and 65 malignant tumors. As a result, the proposed system achieved sensitivities of 100% (133/133), 90% (121/133), and 80% (107/133) with FPs/pass of 9.44, 5.42, and 3.33, respectively. The figure of merit (FOM) of the combination of three feature sets is 0.46 which is significantly better than other feature sets (p-value<0.05). Summarily, the proposed CADe system based on the multidimensional feature sets extracted from the watershed segmentation is promising in detecting tumors in ABUS images.

並列關鍵字

Breast cancer ； Computer-aided detection ； Automated whole breast ultrasound ； Watershed segmentation ； Multi-dimensional false positive reduction

參考文獻

Reference

[1] R. Siegel, D. Naishadham, and A. Jemal, "Cancer statistics, 2013," CA: a cancer journal for clinicians, vol. 63, pp. 11-30, 2013.

[2] A. Jemal, F. Bray, M. M. Center, J. Ferlay, E. Ward, and D. Forman, "Global cancer statistics," CA: a cancer journal for clinicians, vol. 61, pp. 69-90, 2011.

[3] E. A. Sickles, "Breast Imaging: From 1965 to the Present1," Radiology, vol. 215, pp. 1-16, 2000.

[4] M. A. Roubidoux, M. A. Helvie, N. E. Lai, and C. Paramagul, "Bilateral breast cancer: early detection with mammography," Radiology, vol. 196, pp. 427-431, 1995.