Evaluation of an automated breast volume scanner according to the fifth edition of BI-RADS for breast ultrasound compared with hand-held ultrasound

Abstract

ObjectivesTo investigate the automated breast volume scanner (ABVS) in comparison with hand-held ultrasound (HHUS) according to the fifth edition of BI-RADS ultrasound. Material and methodsA total of 831 lesions in 786 patients who underwent both HHUS and ABVS were included. Three radiologists independently evaluated the sonographic features of each lesion according to the fifth BI-RADS edition. The kappa coefficient (κ) was calculated for each BI-RADS descriptor and final assessment category. The accuracy of malignancy prediction and diagnostic performance of the BI-RADS descriptors were assessed using multivariate logistic regression and area under the receiver operator characteristic curve (AUC), respectively. ResultsABVS and HHUS showed moderate to good interobserver agreement (κ = 0.53–0.67 and 0.55–0.70, respectively) except in associated features (κ = 0.31 and 0.36, respectively) for BI-RADS lexicons. Irregular shape, a non-circumscribed margin, and posterior features (combined or shadowing) were independently associated with malignancy in both ABVS and HHUS. Calcification presence on ABVS (odds ratio [OR], 95% confidence interval [CI]: 2.09, 1.11–3.94) and non-parallel orientation on HHUS (OR, 95% CI: 2.04, 1.10–3.78) were independently associated with malignancy. There were no significant differences between ABVS and HHUS in sensitivity (84.2% vs. 84.2%), specificity (80.5% vs. 83.9%), or AUC (0.88 vs. 0.90). ConclusionsAccording to the fifth BI-RADS edition, ABVS is not statistically significantly different from HHUS with regard to interobserver variability and diagnostic performance.