Super‐Resolution Ultrasound for Lymph Node Evaluation: Diagnostic Value of Static Time‐Point Parameters Versus Dynamic Temporal Changes
Qing Xie, Ruixue Cui, Zhendan Zhao, Ming Li, Yuehua Gao, Huawei ZhangObjectives
To evaluate quantitative super‐resolution ultrasound (SRUS) parameters (static central‐to‐peripheral ratios and dynamic temporal changes) for differentiating benign and metastatic lymph nodes, and to compare the performance of corresponding predictive models.
Methods
A prospective study included 172 pathologically confirmed lymph nodes (100 benign, 72 metastatic). SRUS was performed to derive density and velocity parameters, with central‐to‐peripheral ratios and temporal changes calculated. Static and dynamic models were constructed using univariate analysis, least absolute shrinkage and selection operator (LASSO), and logistic regression. Model performance was assessed by cross‐validation, area under the receiver operating characteristic curve (AUC), calibration, decision curve analysis, and clinical impact curve.
Results
Metastatic lymph nodes showed significantly lower central‐to‐peripheral ratios of density and velocity parameters but higher relative temporal changes in density SD and velocity SD compared with benign nodes (all P < .05). The static time‐point model retained 2 SRUS variables (Mean.Density_C_P_T2, Max.Density_C_P_T3), yielding a full‐sample AUC of 0.8607. The temporal change model included 2 dynamic parameters (Velocity SD change from T1 to T2, Mean density change from T2 to T3) and achieved a higher AUC of 0.8715, with better accuracy, specificity, and model fit. Both models showed favorable calibration and clinical net benefit.
Conclusion
Quantitative SRUS parameters effectively reflect pathological angiogenesis in metastatic lymph nodes. The temporal change model provides superior diagnostic performance over the static time‐point model and may serve as a reliable noninvasive tool for distinguishing benign from metastatic lymph nodes.