基于两阶段集成SVM的前列腺肿瘤识别

周涛; 陆惠玲; 陈志强; 马苗

doi:10.3788/OPE.20132108.2137

您当前的位置：

首页 >

文章列表页 >

基于两阶段集成SVM的前列腺肿瘤识别

信息科学 | 更新时间：2020-08-12

- 基于两阶段集成SVM的前列腺肿瘤识别
- Prostate Tumor Recognition Based on Two-stage Ensemble SVM
- 光学精密工程 2013年21卷第8期页码：2137-2145
- 作者机构：
  
  1. 宁夏医科大学理学院,宁夏银川,750004
  2. 宁夏医科大学附属总医院放射科,宁夏银川,750004
  3. 陕西师范大学计算机科学学院,陕西西安,710062
- 作者简介：
- 基金信息：
  
  多源医学图像融合与识别研究();多模态医学图像中ROI区域的识别研究();基于粗糙集的多源医学图像融合与识别研究()
- DOI：10.3788/OPE.20132108.2137
  中图分类号： R445.2;TP391.4
- 收稿日期：2012-12-24，
  
  修回日期：2013-03-05，
  
  网络出版日期：2013-08-20，
  
  纸质出版日期：2013-08-15
- 稿件说明：
移动端阅览
周涛陆惠玲陈志强马苗. 基于两阶段集成SVM的前列腺肿瘤识别[J]. 光学精密工程, 2013,21(8): 2137-2145

ZHOU Tao LU Hui-ling CHEN Zhi-qiang MA Miao. Prostate Tumor Recognition Based on Two-stage Ensemble SVM[J]. Editorial Office of Optics and Precision Engineering, 2013,21(8): 2137-2145
周涛陆惠玲陈志强马苗. 基于两阶段集成SVM的前列腺肿瘤识别[J]. 光学精密工程, 2013,21(8): 2137-2145 DOI： 10.3788/OPE.20132108.2137.

ZHOU Tao LU Hui-ling CHEN Zhi-qiang MA Miao. Prostate Tumor Recognition Based on Two-stage Ensemble SVM[J]. Editorial Office of Optics and Precision Engineering, 2013,21(8): 2137-2145 DOI： 10.3788/OPE.20132108.2137.

摘要

从核磁共振成像（MRI）对前列腺肿瘤的诊断入手，提出了一种基于两阶段集成支持向量机（SVM）的前列腺肿瘤辅助诊断方法。首先，提取MRI图像中的前列腺感兴趣区域（ROI）的统计特征、纹理特征和不变矩特征；然后，在不同的特征空间里，使用不同的核函数来扰动SVM参数并在不同的特征空间生成个体SVM，通过相对多数投票进行第一次集成；接着把第一次集成结果用相对多数投票进行第二次集成；最后，以前列腺患者的MRI图像为原始数据，采用两阶段融合集成SVM对前列腺肿瘤进行辅助诊断。实验显示，第一次集成分类准确率最高比单SVM提高了26.67%，第二次集成分类准确率比第一次集成SVM提高了3.33%，结果表明本文算法能够有效提高前列腺肿瘤的识别精度。

Abstract

On the basis of prostate tumor diagnosis by nuclear Magnetic Resonance Imaging(MRI)

a two-stage integrating Support Vector Machine(SVM) method were proposed to realize the prostate tumor aided diagnosis. Firstly

the statistical features

invariant moment features and the texture feature of the Area of Interest( ROI )for the prostate in a MRI image were extracted. Then

SVM parameters were disturbed by using different kernel functions in different feature spaces

and the first integration was carried out by relative majority voting. Furthermore

the results of first integrating were integrated again by the relative majority voting. Finally

MRI images of prostate patients were regarded as original data

and two-stage integrating SVM were utilized to aid tumor diagnosis. Experiment results show that the classification accuracy from the first integration has improved by 26.67% as compared with that of single-stage SVM and that from the second integration has improved 3.33% than that of the first integration. These results illustrate that the proposed algorithm can improve the recognition accuracy of prostate tumor effectively.

关键词

Keywords

references

何元烈.多模医学图像配准与融合技术及医学智能辅助诊断系统研究[D].广州:华南理工大学，2006.HE Y L.Multimodal medical image registration, fusion and intelligent diagnosis system research[D].GuangZhou:South China University of Technology,2006.(in Chinese)[2]施发表,韦嘉瑚,王占立，等.正常和良性增生前列腺的高场强磁共振成像特征[J].中华老年医学杂志,1997，16(2):79-83.SHI F B, WEI J H, WANG Z L, et al.Normal prostate and benign prostatic hyperplasia:MR imaging at 1.5T[J].Chinese Journal of Geriatrics, 1997,16(2):79-83.(in Chinese)[3]华立新,居小兵,吴宏飞,等.良性前列腺增生和前列腺癌组织中雄激素受体的表达[J].中华老年医学杂志, 2003,22(7):405-408.HUA L X,JU X B,WU H F,et al..The expression of androgen receptor gene in benign prostatic hyperplasia and prostatic cancer tissues [J].Chinese Journal of Geriatrics,2003,22(7):405-408. (in Chinese)[4]NIAF E,ROUVIRE O,MGE-LECHEVALLIER F, et al..Computer-aided diagnosis of prostate cancer in the peripheral zone using multiparametric MRI [J].Physics in Medicine and Biology,2012,57(12):3833-3851.[5]LIOBET R,TOSELLI A H,JUAN C, et al..Computer-aided prostate cancer detection in ultrasonographic images [C].Pattern Recognition and Image Analysis,2003,2652:411-419.[6]TABESH A,TEVEROVSKIY M,PANG H Y,et al..Multifeature prostate cancer diagnosis and gleason grading of Histological [J].IEEE Transactions on Medical Imaging,2007,26(10):1366-1378.[7]钟桦,杨晓鸣,焦李成.基于多分辨共生矩阵的纹理图像分类[J].计算机研究与发展,2011,48(11):1991-1999.ZHONG H,YANG X M,JIAO L CH. Texture classification based on multiresolution co-occurrence matrix [J].Journal of Computer Research and Development,2011,48(11):1991-1999.(in Chinese)[8]薄华,马缚龙,焦李成.图像纹理的灰度共生矩阵计算问题的分析[J].电子学报, 2006,34(1):155-159.BO H,MA F L,JIAO L C. Research on computation of GLCM of image texture [J].Acta Electronica Sinica, 2006,34(1):155-159.(in Chinese)[9]HU M K.Visual pattern recognition by moment invariants[J].IEEE Transactions on Information Theory,1962,8(2):179-187.[10]CARUANA R,NICULESCU-MIZIL A,CREW G,et al..Ensemble selection from libraries of Models[C].Proceedings of the 21th International Conference on Maching Learning,2004,137-144.[11]SAIN S R. The Nature of Statistical Learning Theory [M].New York: Springer-Verlag ,1995.[12]张仁杰,庄松林,臧道青.基于聚类分析与支持向量机模型的缸盖座圈图像判别[J].光学精密工程,2011,19(10):2478-2483.ZHANG R J,ZHUANG S L,ZANG D Q. Estimation of gap width of cylinder socket sleeve based on clustering analysis and SVM [J].Opt. Precision Eng.,2011,19(10):2478-2483.(in Chinese)[13]张春霞,张讲社.选择性集成学习算法综述[J].计算机学报,2011,34(8):1399-1410.ZHANG CH X,ZHANG J SH.A survey of selective ensemble learning algorithms[J].Chinese Journal of Computers, 2011,34(8):1399-1410.(in Chinese)[14]KIM H C, PANG S N, JE H M, et al.. Constructing Support Vector Machine Ensemble [J].Pattern Recognition, 2003,36(12):2757-2767.[15]VALENTINI G, MUSELI M, RUFFINO F. Cancer recognition with bagged ensembles of support vector machines[J].Neuro computing,2004,56:461-466.[16]DUIN R P W.The combining classifier:to train or not to train [C].16th International Conference on Pattern Recognition,2002,2:765-770.

浏览量

109

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据