Value of Multimodal Ultrasound Imaging With Super‐Resolution in the Differential Diagnosis of Benign and Malignant Superficial Lymph Nodes

Objectives Accurate differential diagnosis of benign and malignant superficial lymph nodes (LNs) is critical for clinical management. The objectives of this study were to explore the feasibility and clinical value of super‐resolution ultrasound (SRUS) in the differential diagnosis of superficial lymphadenopathy and to establish and validate a multimodal ultrasound diagnostic model based on conventional ultrasound (US), contrast‐enhanced ultrasound (CEUS), and SRUS, so as to improve the accuracy of distinguishing benign from malignant superficial LNs in clinical practice. Methods A total of 200 LNs from 165 patients with superficial lymphadenectasis were prospectively enrolled from two medical centers. All patients underwent conventional US, CEUS, and SRUS imaging. Predictive models were constructed based on the conventional US, CEUS, and SRUS features in the training cohort, and the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were calculated to evaluate their predictive performance in the external validation cohort. Results The training cohort comprised 117 patients with a total of 140 LNs including 46 benign and 94 malignant lesions. The validation cohort included 48 patients with 60 LNs in total, with 21 benign and 39 malignant lesions. Among the models constructed based on the training cohort, the SRUS model exhibited significantly superior diagnostic performance (AUC = 0.822) to the conventional US and CEUS models ( P < .05). The combined model, integrating all three modalities, achieved an AUC of 0.892 (95% confidence interval [CI]: 0.833–0.952) in the training cohort and 0.890 (95% CI: 0.791–0.990) in the external validation cohort, demonstrating excellent and stable diagnostic performance. Conclusion SRUS provides crucial imaging evidence for the non‐invasive qualitative evaluation of LNs. The multimodal diagnostic model incorporating the features of US, CEUS, and SRUS can significantly improve the accuracy of distinguishing benign from malignant LNs.