Predicting axillary lymph node metastasis in breast cancer via quantitative analysis of VueBox ultrasound contrast-enhanced imaging combined with multimodal feature fusion.

Background: Axillary lymph node metastasis (ALNM) is a major determinant of prognosis and treatment strategy in patients with breast cancer, yet reliable preoperative assessment remains challenging. Conventional ultrasound evaluation is primarily morphology-based and subject to interobserver variability. Contrast-enhanced ultrasound (CEUS) with VueBox enables the quantitative and reproducible assessment of tumor perfusion. This study aimed to develop and validate a nomogram integrating quantitative CEUS parameters with multimodal ultrasound features and serological, pathological, and immunohistochemical indicators for the preoperative prediction of ALNM. Methods: This retrospective study included 82 patients with breast cancer pathologically confirmed between January 2023 and July 2025, comprising 84 breast lesions. According to postoperative histopathological findings, patients were categorized into an ALNM-positive group (n=36) or an ALNM-negative group (n=48). Quantitative analysis of CEUS time-intensity curves (TICs) was performed with VueBox multimodal ultrasound features-including conventional two-dimensional ultrasound, color Doppler flow imaging (CDFI), shear wave elastography (SWE), and qualitative and quantitative CEUS characteristics-which were integrated with serological, pathological, and immunohistochemical indicators. Univariate and multivariate logistic regression analyses were performed to identify independent predictors. A diagnostic nomogram was constructed and evaluated with receiver operating characteristic (ROC) analysis, calibration curves, and decision curve analysis (DCA). Internal validation was conducted via bootstrap resampling with 1,000 iterations. Results: Multivariate logistic regression analysis identified four independent predictors of ALNM: maximum Young modulus [Emax; odds ratio (OR) =1.018; 95% confidence interval (CI): 1.006-1.030; P=0.004], enlarged enhancement scope (OR =10.868; 95% CI: 2.191-53.906; P=0.004), microcalcifications (OR =8.536; 95% CI: 2.092-34.832; P=0.003), and time to peak intensity (TTP; OR =0.849; 95% CI: 0.722-0.998; P=0.047). The nomogram incorporating these variables demonstrated robust discriminatory performance, with an area under the curve (AUC) of 0.926 (95% CI: 0.872-0.980), a sensitivity of 97%, and a specificity of 75%. Internal validation with bootstrap resampling yielded a concordance index (C-index) of 0.926 (95% CI: 0.865-0.974), indicating robust model stability. The calibration curve showed satisfactory agreement between predicted and observed probabilities, and the Hosmer-Lemeshow test indicated good model fit (χ Conclusions: The column nomogram model developed in this study, based on quantitative analysis of VueBox ultrasound contrast imaging and multimodal feature fusion, demonstrated excellent performance in predicting ALNM and holds clinical application value.