乳腺X线图像结构扭曲检测的一种新方法

乳腺X线图像中结构扭曲的检测是乳腺癌诊断的重要手段之一.针对结构扭曲检测假阳性高的问题,运用系统工程原理, 本文提出了一种新的结构扭曲检测方法.该方法通过计算毛刺相似度收敛指数增强放射状毛刺,并利用三环区域模板和收敛门限得到可疑区域,最后使用支持向量机对可疑区域分类.该方法在Mini-MIAS (mammographic image analysis society)图像数据库的64幅乳腺图像和北京大学人民医院乳腺中心50幅临床乳腺图像上进行了验证,实验结果表明, 本文提出的方法能有效地降低假阳性,并且同时适用于东西方女性乳腺图像.

A novel method for detection of architectural distortion in mammograms

Detection of architectural distortion (AD) in mammograms is one of important approaches in breast cancer diagnosis. For solving the problem of high false positives in the detection of AD, a novel method is proposed based on the methodology of systems engineering. In the proposed method, the spiculation similarity convergence index (SSCI) is computed to enhance radiating spiculations. After that, the three-ring-region temple and the convergence threshold are used to obtain suspicious regions. Finally, support vector machine (SVM) is utilized to classify the suspicious regions into AD and normal tissues. The proposed method is evaluated on 64 mammograms from the mammographic image analysis society (Mini-MIAS) and 50 mammograms from the breast cancer of Peking University People's Hospital. The results show that the proposed detection method can decrease FPs significantly, and can be applied to mammograms of both oriental and occidental women.