QUANTITATIVE ULTRASOUND ANALYSIS FOR CLASSIFICATION OF CYST AND SOLID BENIGN BREAST ABNORMALITIES

Abstract

Ultrasound is an important adjunct to mammography in detection and characterization of breast lesions especially in mammographically dense breasts. It has been especially useful in distinguishing cysts from solid tumors. This study was aimed to determine Q values and to classify benign breast abnormalities in ultrasound images based on the Q values. QLAB region of interest (ROI) provides time-intensity curves (data sets) from multiple ROIs applied to the images. Mean intensity of the selected ROIs are known as Q-values. Five established criteria namely outline, perinodular cuffing sign, posterior shadowing, edge shadowing and color flow, based on shape, texture and region characteristics of the breast masses were extracted for cyst and solid classification. The results show that Q-values covering range from 2.61 to 25.81 for cyst and from 19.82 to 77.24 for solid. In conclusion, the Q-values range is defined whose numerical values will help to classify between cyst and solid benign breast abnormalities.