Ultrasound Inversion Imaging Improves the Specificity of BI-RADS Category 4A Masses, Especially in Masses Without Calcification. A Single-Center Experience.

Abstract

Objective: This study aimed to assess whether ultrasound inversion imaging enhances the specificity of Breast Imaging Reporting and Data System (BI-RADS) BI-RADS 4A masses, reducing unnecessary biopsies.

Method: From July 2021 to March 2023, 425 pathologically confirmed breast masses were collected from the Second Affiliated Hospital of Anhui Medical University. Based on a five-point scale of inversion imaging criteria proposed in a previous study, the diagnostic performance of three classification algorithms, including BI-RADS, BI-RADS category after inversion imaging, and BI-RADS category after inversion imaging without calcification masses-via receiver operating characteristic (ROC) curve analysis.

Results: Among the 425 breast masses, 116 masses were classified as BI-RADS category 4A (82 benign, 34 malignant). After inversion imaging, 74 benign BI-RADS category 4A masses were correctly downgraded to BI-RADS category 3, avoiding biopsies. ROC curve values indicated that the specificity, sensitivity, and accuracy were 31.06%, 98.48%, and 72.94%, respectively, with BI-RADS, 72.05%, 95.83%, and 86.82%, respectively, with BI-RADS after inversion imaging, and 86.51%, 97.75%, and 93.09%, respectively, with BI-RADS category after inversion imaging without calcification masses.

Conclusion: Ultrasound inversion imaging offers a feasible supplementary method to distinguish between benign and malignant breast masses, improving BI-RADS 4A specificity, especially for non-calcified masses.