IMPROVING THE DISTINCTION BETWEEN BENIGN AND MALIGNANT BREAST-LESIONS- THE VALUE OF SONOGRAPHIC TEXTURE ANALYSIS

To improve the ability of ultrasound to distinguish benign from malign ant breast lesions, we used quantitative analysis of ultrasound image texture. Eight cancers, 22 cysts, 28 fibroadenomata, and 22 fibrocysti c nodules were studied. The true nature of each lesion was determined by aspiration (for some cysts) or by open biopsy. Analysis of image te xture was performed on digitized video output from the ultrasound scan ner using fractal analysis and statistical texture analysis methods. T he most useful features were those derived from co-occurrence matrices of the images. Using two features together (contrast of a co-occurren ce matrix taken in an oblique direction, and correlation of a co-occur rence matrix taken in the horizontal direction), it was possible to ex clude 78% of fibroadenomata, 73% of cysts, and 91% of fibrocystic nodu les while maintaining 100% sensitivity for cancer. These findings sugg est that ultrasonic image texture analysis is a simple way to markedly reduce the number of benign lesion biopsies without missing additiona l cancers. (C) 1993 Academic Press, Inc.