张慧,钱林学,刘红艳.智能乳腺全容积超声鉴别诊断肉芽肿性乳腺炎与浸润性导管癌[J].中国医学影像技术,2019,35(8):1210~1213
智能乳腺全容积超声鉴别诊断肉芽肿性乳腺炎与浸润性导管癌
Intelligent breast full volume ultrasound in differential diagnosis of granulomatous mastitis and invasive ductal carcinoma
投稿时间:2019-01-14  修订日期:2019-06-06
DOI:10.13929/j.1003-3289.201901060
中文关键词:  乳腺炎  癌,导管,乳腺  超声检查
英文关键词:mastitis  carcinoma,ductal,breast  ultrasonography
基金项目:北京市医院管理局"登峰"人才培养计划(DFL20180102)。
作者单位E-mail
张慧 首都医科大学附属北京友谊医院超声科, 北京 100050  
钱林学 首都医科大学附属北京友谊医院超声科, 北京 100050 qianlinxue2002@163.com 
刘红艳 首都医科大学附属北京友谊医院超声科, 北京 100050  
摘要点击次数: 2654
全文下载次数: 874
中文摘要:
      目的 探讨智能乳腺全容积超声系统(IBUS)鉴别诊断肉芽肿性乳腺炎(GM)与浸润性导管癌(IDC)的价值。方法 回顾性分析经病理证实的36例GM和76例IDC患者(实验组)的IBUS影像学资料,建立多因素Logistic回归模型,筛选出主要超声特征,并评价其鉴别诊断GM与IDC的价值。使用逻辑回归公式计算概率,评价Logistic回归模型在验证组(36例GM、59例IDC)中的诊断效能。结果 多因素Logistic回归筛选出5个鉴别诊断GM与IDC的独立危险因素,依次为"牵拉聚集"征(优势比44.33)、钙化(优势比23.08)、周边晕环(优势比20.35)、导管扩张(优势比19.01)和"火山口"征(优势比14.40),其中"牵拉聚集"征是优势比最大的独立危险因素,鉴别诊断GM和IDC最有价值。IBUS诊断GM的准确率为91.67%(33/36),诊断IDC的准确率为89.83%(53/59)。结论 IBUS鉴别诊断GM与IDC具有较高价值。
英文摘要:
      Objective To explore the value of intelligent breast full volume ultrasound system (IBUS) in differential diagnosis of granulomatous mastitis (GM) and invasive ductal carcinoma (IDC). Methods Ultrasonographic data of 36 GM patients and 76 IDC patients were analyzed retrospectively. Multivariate Logistic regression model was established to screen out the main ultrasonic features. The value of differential diagnosis of GM and IDC, the diagnostic efficiency of Logistic regression model were evaluated, ROC curve was drawn. Results Multivariate Logistic regression screened out 5 independent risk factors for differential diagnosis of GM and IDC, including pull and gather sign (odds ratio=44.33), calcification (OR=23.08), halo ring (OR=20.35), catheter dilatation (OR=19.01) and crater sign (OR=14.40). The accuracy of IBUS in diagnosis of GM and IDC was 91.67%(33/36) and 89.83%(53/59), respectively. Conclusion IBUS has good diagnostic and clinical value in differential diagnosis of GM and IDC.
查看全文  查看/发表评论  下载PDF阅读器