Automatic Breast Lesion Segmentation and Scoring in Color Elastography Images

While breast biopsy remains the gold standard for diagnosis of suspicious lesions, a large proportion of biopsy specimens reveal a benign result. Therefore, a noninvasive and reliable method to identify low-risk lesions will be a valuable tool. Ultrasound has been an important imaging technique for detecting breast tumors. As opposed to the conventional BrightnessMode image, strain elastography is a new technique for imaging the elasticity and is useful in detecting the stiffness of tissues. In this paper, we propose a Computer Aided Diagnosis(CAD) system to automatically segment breast lesions and then extract features for efficient and automatic classification of the tumor into different scores based on Color Elastography Scoring system proposed by Itoh et al.