Breast imaging reporting and data system categorization of breast lesions: Comparison between digital breast tomosynthesis and ultrasonography.

BACKGROUND

Breast cancer is a leading cause of death in women worldwide. Digital Mammography (DM) has long been the cornerstone of breast cancer screening; however, its limitations in dense breast tissue have led to the use of other modalities, such as Digital Breast Tomosynthesis (DBT) and ultrasound (US), which offer unique diagnostic advantages.

OBJECTIVE

To compare the diagnostic performance of DBT and breast US for characterizing breast lesions according to BIRADS categorization.

METHODS

This prospective study was conducted over 18 months at the Mansoura University radiology department. It included 120 female patients (aged 31-74 years) who underwent DBT and breast US. Two experienced breast radiologists independently assessed the lesions; observer A evaluated US features, and observer B evaluated DBT criteria according to the BIRADS Atlas 2013, with subsequent pathological correlation of the examined lesions. DM was done before DBT, but the mammography results were not analyzed or included in this work.

RESULTS

A total of 154 lesions were evaluated. Among them, 64 (41.6%) were benign, and 90 (58.4%) were malignant. The total number of masses detected by DBT was 150, and by US was 152. There was a good to very good agreement between observer A and observer B regarding BIRADS category of the lesions (kappa > 0.8). There was moderate to good agreement between US and DBT on mass margin (k = 0.664). There was a very good agreement between US and DBT on mass shape (k > 0.8). DBT showed higher specificity for lesion shape (100%) and BIRADS classification (90.63%), while US exhibited higher sensitivity for mass margin (95.56%) and BIRADS categorization (93.33%). Significant diagnostic differences were noted between benign and malignant lesions with regard to the lesion-specific imaging features in laterality, lesion morphology, vascularity, echogenicity, and BIRADS score.

CONCLUSION

DBT has diagnostic performance comparable to US in characterizing breast lesions as benign or malignant according to BIRADS categorization and offers the advantage of improved lesion margin detection due to decreased tissue overlap.

PLAIN LANGUAGE SUMMARY: Breast scans help doctors decide if a lump is harmless or cancer. This study compared two types of scans, one using layered X-ray images and one using sound waves, in people with breast lumps. This study found that both methods worked well, but each had strengths for different image features. This matters because using the right scan can help improve diagnosis and support better care.