CAD在乳腺纤维腺瘤钼靶X线诊断中的应用  

The Value of CAD in the Mammography Diagnosis for Breast Fibroadenoma

作　　者：林盼盼 黄恺[3] 吴仁华[3] LIN Pan-pan;HUANG Kai;WU Ren-hua(School of Clinical Medicine,Quanzhou Medical College,Fujian 362000,China;Department of Imaging,People’s Hospital of Quanzhou Medical College,Fujian 362000,China;Department of Radiology,Second Affiliated Hospital,Shantou University School of Medicine,Guangdong 515041,China)

机构地区：[1]泉州医学高等专科学校临床医学院,福建362000 [2]泉州医学高等专科学校附属人民医院影像科,福建362000 [3]汕头大学医学院第二附属医院放射科,广东515041

出　　处：《影像技术》2021年第5期17-21,共5页Image Technology

基　　金：国际(地区)合作与交流项目(82020108016);国家自然科学基金项目(31870981);广东省高水平大学建设计划临床医学重点建设学科专项资金项目(002-18120302);福建省中青年教师教育科研项目(JAT191468)。

摘　　要：目的:探讨计算机辅助检测(CAD)在乳腺纤维腺瘤钼靶X线诊断中的应用价值。方法:收集本院109例乳腺纤维腺瘤患者的钼靶图像,使用CAD系统标识感兴趣区域(ROIs)并观察系统勾画的病灶位置及形状,探讨CAD在乳腺纤维腺瘤诊断中的应用价值。结果:在109例乳腺纤维腺瘤患者钼靶图像中,CAD系统对脂肪型乳腺的识别率最高,达到100%,对致密型、多量腺体型和少量腺体型乳腺的识别率分别为5.88%、5.41%和4.55%。结论:CAD能够识别高于其周围腺体组织密度的乳腺纤维腺瘤,本结论对于数字图像处理的优化研究有一定参考价值。Objective:To investigate the value of computer-aided detection(CAD)in the diagnosis of mammography for breast fibroadenoma.Methods:We collected 109 breast fibroadenoma patients’mammogram images,identifiedthe regions of interest(ROIs)by using CAD software,and observed the breast lumps’location and shape.Results:Among the mammograms of 109 breast fibroadenoma patients,the CAD software has the 100%recognition rate for fatty breasts,and 5.88%,5.41%and 4.55%for dense breasts,multiple glandular breastsand small amounts of glandular breasts,respectively.Conclusions:CAD could identify breast fibroadenoma with higher density than its surrounding glandular tissue,and the conclusion providescertain reference values for the optimization of image processing software.

关 键 词：乳腺纤维腺瘤 钼靶 计算机辅助检测 

分 类 号：R445.4[医药卫生—影像医学与核医学] R814[医药卫生—诊断学]