【摘要】 目的：探讨超声人工智能(AI)联合剪切波弹性成像(SWE)在乳腺结节良恶性鉴别诊断中的临床效能。方法：回顾性分析200例乳腺良恶性肿瘤患者(共206个结节)的数据，所有病灶均接受SWE及超声AI检查。结果：经病理组织学检查确诊恶性肿瘤98个(47.57%),良性肿瘤108个(52.43%)。乳腺恶性病灶的弹性模量相关指标及AI辅助评分均显著高于良性病灶(P<0.05)。ROC曲线分析显示，SWE与超声AI在乳腺结节良恶性鉴别诊断中均有效，且二者联合使用的效能优于单一指标(P<0.05)。结论：超声AI与SWE联合使用可提高乳腺结节良恶性的鉴别诊断效能。更多还原

【Abstract】 Objective: To investigate the clinical efficacy of ultrasonic artificial intelligence(AI) combined with shear wave elasto-graphy(SWE) in the differential diagnosis of benign and malignant breast nodules, aiming to provide more references for enhancing the efficiency of ultrasonic differential diagnosis. Methods: The data of 200 patients with benign and malignant breast tumors(totaling 206 nodules) were retrospectively analyzed. All lesions were examined by SWE and ultrasonic AI examinations. Results: Pathological examination confirmed 98 malignant tumors(47.57%) and 108 benign tumors(52.43%). The elastic modulus-related indices and AI-assisted scores of malignant breast lesions were significantly higher than benign breast lesions(P<0.05). ROC curve analysis demonstrated that both SWE metrics(including maximum elastic modulus, average, and standard deviation) and AI-assisted scores were effective in differentiating benign from malignant breast nodules, with the combination of both showing superior performance compared to individual metrics(P<0.05). Conclusion: The combined application of ultrasonic AI and SWE can improve the efficiency of differential diagnosis of benign and malignant breast nodules.更多还原