Breast MRI: Simplifying protocol and BI-RADS categories

Abstract

Background

Breast magnetic resonance imaging (MRI) is a useful imaging method, and studies have proven abbreviated protocols may reduce its time and costs with comparable accuracy. To assess the agreement of the abbreviated and full MRI protocols for breast cancer.

Patients and Methods

This retrospective cross-sectional study included consecutive women who underwent breast MRI from June 2019 to May 2020. Breast MRIs were performed in 2 hospitals, and 3 board-certified radiologists conducted its evaluations. The abbreviated breast MRI was assessed in the following order: MIP; subtraction; and T1W pre- and post-contrast. Findings were categorized according Breast Imaging - Reporting and Data System (BI-RADS 5th edition). The groups were named simplified BI-RADS and were adapted from BI-RADS 5-point scoring system.

Results

Four hundred and nineteen women (838 BI-RADS), with mean age of 49.5 ± 12.3 years were evaluated. A very good agreement was found between the protocols for total population (weighted kappa: 0.95; P <.001) and for MRI indication subgroups (weighted kappa: 0.86-1.00; P <.001). A high accuracy was also observed comparing abbreviated breast MRI compared to a full diagnostic MRI (97.6%).

Conclusion

Abbreviated breast MRI has demonstrated performance comparable to full diagnostic MRI protocols for breast cancer screening and diagnostic. In addition, the present study proposes the use of the simplified BI-RADS as a new classification, in order to assist physicians in clinical management.