A multicenter reader study designed to examine the effect of CAD on the diagnostic accuracy of breast MRI indicated improved sensitivity and comparable accuracy among novice and expert readers reviewing images without and with CAD, according to a study published in the June issue of American Journal of Roentgenology. CAD software did not increase interpretation time.

Although breast MRI CAD is widely used to help radiologists manage the hundreds of datasets generated by an exam, few studies have assessed diagnostic accuracy and interpretation time. Thus, Constance D. Lehman, MD, from the section of breast imaging at University of Washington, Seattle, and colleagues, designed a study to compare diagnostic accuracy among novice and expert readers reviewing breast MRI images without and with CAD.

Nine expert readers, defined as imagers who had interpreted 100 breast MRI exams prior to the study, and 11 novice readers participated in two image review sessions spaced six months apart to minimize recall bias. The case series included 70 breast MRI exams, 27 benign and 43 malignant.

The researchers used three scales to assess each reader’s interpretation: modified BI-RADS, probability of malignancy scale and percentage probability of malignancy scale. “Results across interpretative scales were remarkably consistent,” wrote Lehman et al.

Sensitivity for experienced readers improved area under the receiver operating characteristic curve (AUC) from 0.84 to 0.91. Novice readers bumped from 0.77 to 0.83.

“Diagnostic accuracy, measured by AUC, for novices without CAD was 0.77, for novices with CAD was 0.79, for experienced readers without CAD was 0.80, and for experienced readers with CAD was 0.83.” The upward trend was not statistically significant.

Interpretation time held steady without and with CAD.

Lehman and colleagues calculated reader-specific AUCs, which showed reader-to-reader variability in the effect of CAD on diagnostic accuracy. Further dissemination of CAD in practice requires additional focus on the relationship between CAD and diagnostic accuracy, according to the researchers.

The researchers concluded by calling for education programs to focus on improving diagnostic accuracy and efficiency of breast MRI CAD.