Age-integrated breast imaging reporting and data system assessment model to improve the accuracy of breast cancer diagnosis

Purpose: Ultrasonography is one of the most effective methods for diagnosing breast diseases, especially among Asian women. The Breast Imaging Reporting and Data System (BI-RADS) is widely used independently of age in diagnosing breast cancer via ultrasonography. This study aimed to develop a model that integrates age with the BI-RADS lexicon to improve the diagnostic accuracy of ultrasonography in diagnosing breast cancer among the Chinese population. Methods This study comprised two cohorts: the training cohort, including 975 women from the Renmin Hospital of Wuhan University, and the validation cohort, including 500 women from the Maternal and Child Health Hospital of Hubei Province. Logistic regression was used to construct a model combining BI-RADS scores with age and to determine the age-based prevalence of breast cancer to predict a cut-off age. The area under the curve (AUC) was used to determine the model’s diagnostic efficacy. Results The age with BI-RADS scores model had the best performance compared to the age-only model and BI-RADS scores-only model with an AUC of 0.872 (95% CI: 0.850–0.894, p < 0.001). Moreover, among participants aged < 30 years, the prevalence of breast cancer was lower than the lower limit of the reference range (2%) for the BI-RADS subcategory 4A lesions but within the reference range for BI-RADS category 3 lesions according to the linear regression analysis. Conclusions The integrated assessment model based on age and BI-RADS may improve the accuracy of ultrasonography in diagnosing breast lesions. Young patients with BI-RADS subcategory 4A lesions may be excluded from biopsy.