Accuracy of classification of breast ultrasound findings based on criteria used for BI-RADS.

Abstract

To assess the accuracy of categorization of breast ultrasound findings based on scoring for malignancy using the sonographic breast imaging-reporting and data system (BI-RADS). Breast ultrasound was performed in 2462 patients between 2001 and 2004 at our unit. Sonographic findings were scored using analog criteria as in BI-RADS for breast ultrasound (mass shape, margin, orientation, posterior acoustic features, lesion boundary, echo pattern). Each lesion was described using these features and classified into categories 1 to 5 according to the BI-RADS for breast ultrasound. Categorization and biopsy results were compared. In twenty-two (0.9%) patients breast ultrasound could not be evaluated because of extreme density of tissue. Normal breast ultrasound belonging to Category 1 was found in 871 (35.4%) patients. Simple cysts classified as Category 2 were observed in 712 (28.9%) women. In 491 (19.9%) patients, apparently benign solid masses (Category 3) were found. Suspicious masses were observed in 225 (9.1%) women and masses highly suggestive of malignancy were found in 141 (5.7%) patients (Categories 4 and 5, respectively). Histological examinations were available from 84 (17.1%) masses that had been classified by BI-RADS as Category 3, in 97 (43.1%) from Category 4 and 106 (75.2%) from Category 5. Accordingly, the rate of malignant findings was 1.2% (n = 1) in Category 3, 17% (n = 16) in Category 4 and 94% (n = 100) in Category 5. Scoring breast ultrasound findings for malignancy based on criteria used for BI-RADS breast ultrasound has a high accuracy, comparable to that obtained by BI-RADS for mammography.