A team led by Dr. Jan van Zelst of Radboud University Medical Center in Nijmegen, the Netherlands, investigated whether dedicated CAD software would enhance radiologists' ABUS interpretations and reduce reading time. The researchers included 120 ABUS exams in the study (30 malignant, 30 benign, and 60 normal). All the cases had histological data or more than two years of negative follow-up that could be used to confirm interpretations.
The exams were read by eight dedicated breast radiologists, once without using CAD and once with the software. Reading sessions were at least eight weeks apart, according to van Zelst and colleagues.
What did the study show? The mean area under the curve (AUC) was comparable for both modes, with conventional ABUS at 0.82 and ABUS with CAD at 0.83. Using CAD improved specificity for seven of the eight readers, while sensitivity decreased somewhat for two of the eight. Mean reading time decreased from 2.56 minutes to 2.22 minutes.
The data suggest that using dedicated CAD software for ABUS improves its efficiency, the group concluded.
