Abstract:

Objective

To evaluate the diagnostic performance of conventional ultrasound combined with automated breast volume scanning （ABVS） for breast non-mass lesions （NMLs）.

Methods

From January 2023 to November 2023， a total of 152 patients preliminarily diagnosed with breast NMLs by routine ultrasound at the First Affiliated Hospital of Harbin Medical University were retrospectively collected.NMLs were categorized into ductal and non-ductal types. The ultrasonic characteristics， including blood flow distribution， microcalcification， internal echogenicity， posterior features， peripheral structural distortion，axillary lymph node abnormalities， convergence sign， stierlin sign， lesion distribution， ductal morphology changes of ducted NMLs， and directional features of non-ducted NMLs， were systematically recorded using both conventional ultrasound and ABVS. Using pathological results as the gold standard， univariate analysis was conducted based on the ultrasonic findings from both modalities. Significant independent variables were subjected to Logistic stepwise regression analysis to identify independent influencing factors. Based on these findings， a diagnostic model integrating conventional ultrasound with ABVS was established， and its diagnostic efficacy was compared with that of conventional ultrasound alone using the Z-test.

Results

Multivariate logistic regression analysis revealed that microcalcification， convergence sign （ABVS），ductal rigidity， and amputation were independent predictors of benign versus malignant ductal NMLs （odds ratio ［OR］=55.522， 12.321， and 3.740， P=0.001， 0.015， 0.033，respectively）. The sensitivity， specificity，and area under the curve （AUC） for conventional ultrasound alone and the model integrating conventional ultrasound with ABVS in diagnosing benign and malignant ductal NMLs were 0.679， 0.677， 0.678， and 0.786， 0.968， 0.877， respectively. The diagnostic performance of the combination model was significantly higher than that of conventional ultrasound alone （Z=-3.479， P=0.001）. Microcalcification， convergence sign（ABVS）， and blood flow signal were identified as independent predictors of malignant non-ductal NMLs（OR=46.435， 11.914， and 9.594， P=0.002， 0.002， 0.013，respectively）. The sensitivity， specificity， and AUC for conventional ultrasound alone and the model integrating conventional ultrasound with ABVS in differentiating benign from malignant NMLs were 0.867， 0.848， 0.858， and 0.917， 0.909， 0.913， respectively.No statistically significant difference was observed in diagnostic efficacy between the two models （P ＞ 0.05）.

Conclusion

Compared to conventional ultrasound， the model integrating conventional ultrasound with ABVS enhances the diagnostic performance for NMLs. This model demonstrates greater diagnostic value for ductal NMLs but does not provide significant advantages in the qualitative diagnosis of non-ductal NMLs.

Key words: Breast, Non-mass lesions, Ultrasound, Automated breast volume scanning