Comparison between automated breast volume scanner (ABVS) versus hand-held ultrasound as a second look procedure after magnetic resonance imaging

Objectives To evaluate the agreement between automated breast volume scanner (ABVS) and conventional ultrasound (US) as a second-look (SL) tool for assessing additional find- ings found on MRI. Methods Over a 7-month period, we prospectively assigned to SL-US and SL-ABVS all patients undergoing 1.5 T breast MRI in whom additional findings were found. Five experi- enced breast radiologists independently interpreted SL-US and SL-ABVS in blinded sessions to evaluate the detection rate of MRI findings and assign them to BI-RADS categories. We calculated the agreement between the two methods in assessing MRI findings as significant (BI-RADS 3-5) versus not significant (BI-RADS 1-2), as well as their cancer detec- tion rate. Results In a population of 131 patients, SL-ABVS and SL-US showed a comparable detection rate of MRI findings (69.3 vs. 71.5%) (p > 0.05; McNemar test), with an almost perfect agreement in assessing them as significant or not (k = 0.94). This translated into a comparably high cancer detection rate (83.8% for SL-ABVS vs. 87.0% for SL-US). Only 1/31 can- cers was missed by SL-ABVS. Conclusions SL-ABVS and SL-US are nearly equivalent in assessing the significance of MRI findings, leading to a com- parable cancer detection rate. SL-ABVS has the potential to replace SL-US in the SL scenario.