Abstract
OBJECTIVES: To evaluate the effectiveness of breast MRI, including diffusion-weighted imaging (DWI), in detecting residual lesions in patients with malignancy after excisional biopsy. METHODS: From January 2018 to December 2023, 3T breast MRI was performed to assess lesion morphology, residual size, and enhancement kinetics. The apparent diffusion coefficient (ADC) values were measured, and the diagnostic outcomes of CE-MRI, CE-MRI with DWI, mammography (MG), and ultrasound (US) were compared with clinical and histopathological data. RESULTS: A total of 152 lesions were analyzed, with 36.2% showing residual malignancy. Both CE-MRI and CE-MRI with DWI effectively identified residual lesions, with significant differences in morphology, size, kinetic patterns, and ADC values (all p < 0.001). CE-MRI with DWI showed a sensitivity of 90.9% and an NPV of 93.6%, compared with 89.1% sensitivity and 92.2% NPV for CE-MRI alone. Sensitivities for MG and US were 57.1% and 38.7%, with NPVs of 64.7% and 59.6%, respectively. Diagnostic accuracy was highest for CE-MRI with DWI (80.9%), followed by CE-MRI (79.0%), MG (60.3%), and US (59.7%). The AUC for CE-MRI with DWI (0.831) was slightly higher than CE-MRI alone (0.811), though not significant (p = 0.095). AUCs for MG and US were lower at 0.623 and 0.563, with no significant difference between MG and US (p = 0.234). CONCLUSIONS: CE-MRI with DWI and CE-MRI alone were comparable and demonstrated excellent performance in discriminating between women with and without residual disease. Integrating CE-MRI with DWI could become a standard protocol for patients with suspected residual malignancy after excisional biopsy.