Breast cancers initially detected by hand-held ultrasound: detection performance of radiologists using automated breast ultrasound data

Abstract

Ultrasonography (US) has been used as an important adjunct to mammography (MG), and automated breast US (ABUS) scanners were originally designed to effectively examine the breast in its entirety. To retrospectively assess the performance of radiologists in the detection of breast cancers, initially detected by hand-held ultrasound (HHUS), using 3D breast volume data obtained from a commercial ABUS system. Bilateral whole breast US was performed using ABUS in 61 consecutive women who were scheduled to undergo US-guided needle biopsy due to suspicious breast masses detected during screening HHUS. Fourteen cancers in 13 women and 48 normal breasts of 48 women with benign disease in the contralateral breast were selected. Three radiologists who had not performed the HHUS examinations independently reviewed the 3D ABUS data for any lesions that required recall for further evaluation. Sensitivities and false-positive rates were calculated. The sensitivities of the three readers for cancer detection were 78.6% (11/14), 78.6%, and 57.1% (8/14), respectively, with false-positive rates of 20.8% (10/48), 12.5% (6/48) and 8.3% (4/48). Seven cancers were detected by all three readers, four cancers by two readers, and one cancer by one reader. Two invasive cancers were not detected by any reader. Of HHUS-detected cancers, only 57.1-78.6% were identified with ABUS. A substantial level of experience and training is necessary to improve cancer detection by ABUS.