[1]回宋立新‘,李东红裴.一种双尺度模板的双视图乳腺肿块检测匹配[J].哈尔滨理工大学学报,2017,(02):129-134.[doi:10.15938/j.jhust.2017.02.024]
 SONG Li-xinLI DonR--honR-Z,PEI henR-ZENG haNIU Bin.Research of Two一view Mammographic Masses Detectionand Matching Based on Double一Template[J].哈尔滨理工大学学报,2017,(02):129-134.[doi:10.15938/j.jhust.2017.02.024]
点击复制

一种双尺度模板的双视图乳腺肿块检测匹配()
分享到:

《哈尔滨理工大学学报》[ISSN:1007-2683/CN:23-1404/N]

卷:
期数:
2017年02期
页码:
129-134
栏目:
计算机与控制工程
出版日期:
2017-04-25

文章信息/Info

Title:
Research of Two一view Mammographic Masses Detection and Matching Based on Double一Template
文章编号:
1007-2683(2017)02-0129-06
作者:
回 宋立新‘李东红2裴
(1哈尔滨理工大学 电气与电子工程学院,黑龙江哈尔滨150080;2聚束科技(北京)有限公司,北京100176; 3中国兵器工业导航与控制技术研究所,北京100089; 哈尔滨理工大学测控技术与通信工程学院,黑龙江哈尔滨150080)
Author(s):
SONG Li-xin LI DonR--honR-ZPEI henR- ZENG ha矿 NIU Bin4
(1. School of Electrical and Electronic L,n}ineerin} 日arhin I」 nicarsitv of Science and ’fechnolo}y,llarbin 150080,China; 2. Focus e-Beam’fechnolo}y(Beijin})Company Limited,Beijin},100176,China; 3. Navigation and Control’fechnolo}y Research lnstitute of China North lndustries Croup Corporation, Beijing 100089,China; 4. School of Measurement and Control ’fechnolo}y and Communications F,n}ineerin}, llarbin Lniversity of Science&’fechnolo}y, llarbin 150080,China)
关键词:
关键词:双视图模板匹配双尺度Sech模板互信息匹配率
Keywords:
Keywords:double-viewtemplate matchingdouble Seeh-templatescorrelation coefficientmutual informa- ton
DOI:
10.15938/j.jhust.2017.02.024
文献标志码:
A
摘要:
摘要:为减少乳腺肿块检测到假阳性区域,进而提高可疑病灶区域的匹配率,提出了一种基 于双尺度模板检测乳腺肿块可疑病灶区域的方法。该方法首先依据CC视图中可疑病灶区域,在 MLO视图中构建条形匹配区域带;然后,基于双尺度Sech模板对肿块图像进行检测后,再做归一 化互相关计算,检测出相关性高的区域为肿块的可疑病灶区域,依据基于形状、面积特征的规则删 除假阳性区域;最后,根据基于互信息的相似性度量方法实现双视图的可疑病灶区域的匹配。实验 结果显示:对DDSM数据库中已确诊的100幅肿块图像进行实验对比,有90幅图像能实现双视图 肿块的匹配,匹配率达到90%,与基于灰度分层的乳腺肿块的双视图匹配相比,匹配率得到提高。
Abstract:
Abstract:In order to reduce the detection of false positive m月SSeS and improve the matching rate of the double V1e41 lump, a method of double templates matching to detect the suspicious lesions area has been proposed. paper then, uses the susplelOUS leslOns In the CC-V1eW area t0 1(lentlfy the matChlng bar area In the MLO-mew firstly. This fend }t uses double Sech-template to detect lumps,the areas of high correlation coefficient will be suspicious }eS1l1nS }3.re}3.. After deleting the false positive regions,it fuse the different size template results based shape and area. information. This Finally, paper m月tc卜. it is realized that the m月SS matching which have by measuring the similarity based On rules of mutual selects 100 pairs The experimental of images been confirmed to carry out the experiments of suspicious lesions mass matching in results show that ninety pairs of images have been achieved double-view comparing with the gray-scale layering algorithm.

备注/Memo

备注/Memo:
收稿日期:2015一12 - 29 基金项目:黑龙江省自然利一学基金(H’200912);哈尔滨创新人才基金(2010RH XXS026) 作者简介:宋立新(1963-),男,博士,教授,L,-mail;lixins99C yalioo. com. cn; 李东红(1988-),女,硕士研究生; 装恒(1982-) ,男,硕士,工程师.
更新日期/Last Update: 2017-06-13