Conventional ultrasound combined with BI-RADS classification in differential diagnosis of benign and malignant complex cystic and solid breast lesions

To explore the diagnostic and therapeutic value of the ultrasound Breast Imaging Reporting and Data System (BI-RADS) in complex cystic and solid breast lesions in different ultrasound subgroups. Clinical data were retrospectively analyzed from 158 patients with 183 complex cystic and solid breast lesions that underwent ultrasound examination and confirmed by histopathology by the Dongyang Affiliated Hospital at Wenzhou Medical University between January 2020 and March 2023. Lesions were classified into 3 subgroups based on the proportion and distribution of solid-cystic components (Group A: thick outer wall and/or septa; Group B: predominantly cystic; and Group C: predominantly solid). Conventional ultrasound features in each group were analyzed against pathological results, BI-RADS classifications were performed, and receiver operating characteristic (ROC) curves were constructed and calculate the area under the curve (AUC). Of 183 lesions, 26 were in Group A (24 benign vs 2 malignant; AUC = 0.708), 76 in Group B (67 benign and 9 malignant tumors; AUC = 0.864), and 81 in Group C (42 benign vs 39 malignant; AUC = 0.921). The overall AUC for all lesions was 0.906. The age of Group B's malignant lesions and the patient age and maximum diameter of malignant lesions in Group C were substantially greater than those in benign lesions (P < .05). Group A exhibited no significant ultrasonic feature differences between benign/malignant lesions (P > .05). In Group B, malignant lesions differed significantly in margin irregularity, surrounding structural distortion, malignant halo, microcalcification, and increased vascularity (P < .05). In Group C, differences were statistically significant in shape, margin, surrounding structural distortion, malignant halo, microcalcification, vascularity, and abnormal axillary lymph node between benign and malignant lesions (P < .05). It was further discovered that there was no statistical significance in the area under the ROC curve in the comparison between groups by the DeLong test method (P > .05). Ultrasonographic features combined with ultrasound BI-RADS classification assist in differentiating between benign and malignant complex cystic and solid breast lesions across various ultrasound subgroups, with better diagnostic value in Subgroups B and C and relatively lower efficacy in Subgroup A. Our findings may contribute to a reduction in unnecessary biopsies.