基于临床及超声特征预测乳腺良恶性结节列线图模型的建立及应用

Establishment and Application of a Nomograms Model for Predicting Benign and Malignant Breast Nodes Based on Clinical and Ultrasound Characteristics

目的基于临床及超声特征建立乳腺良恶性结节鉴别诊断的预测模型。方法回顾性分析2021年7月—2022年11月广西壮族自治区桂林市中医医院乳腺科收治的186例乳腺结节患者的超声声像图及临床资料,采用单因素及多因素Logistic回归分析,筛选出能预测乳腺良恶性结节的独立危险因素,并构建列线图预测模型。结果年龄、结节边缘、周围组织改变和阻力指数为乳腺恶性结节的独立危险因素(P均<0.05)。基于上述4个影响因素构建的列线图模型的受试者操作特征(re‐ceiver operating characteristic,ROC)曲线下面积为0.954(95%CI:0.927~0.981),灵敏度为0.947,特异度为0.901。结论基于年龄、边缘、周围组织改变和阻力指数建立的列线图模型可为鉴别乳腺良恶性结节提供重要依据。

Objective To establish a prediction model for differential diagnosis of benign and malignant breast nodules based on clinical and ultrasound features.Methods Ultrasound sonograms and clinical data of 186 cases of breast nodules in the Department of Mammography of Guangxi Zhuang Autonomous Region Guilin Hospital of Traditional Chinese Medicine from July 2021 to November 2022 were retrospectively selected.Unifactorial and multifactorial logistic regression analyses were used to screen the independent risk factors that could predict benign and malignant breast nodules and construct a nomograms prediction model.Results Age,margins,surrounding tissue changes,and resistance index were independent risk factors for breast malignant nodules(all P<0.05).The area under the receiver operating characteristic(ROC)curve was 0.954(95%CI:0.927-0.981),with a sensitivity of 0.947 and a specificity of 0.901 for the nomograms model constructed based on the above four influencing factors.Conclusion The nomograms model based on age,margin,surrounding tissue changes and resistance index can provide an important basis for identifying benign and malignant breast nodules.