|本期目录/Table of Contents|

[1]刘银凤△,张俊杰,周涛,等.基于多维特征和支持向量机核函数优化的自动化肺结节检测模型[J].生物医学工程研究,2016,02:75-80.
 LIU Yinfeng,ZHANG Junjie,ZHOU Tao,et al.An Automated Lung Nodules based on Multidimensional Characteristics and the Support Vector Machine(SVM) Kernel Function Optimization[J].Journal of Biomedical Engineering Research,2016,02:75-80.
点击复制

基于多维特征和支持向量机核函数优化的自动化肺结节检测模型(PDF)

《生物医学工程研究》[ISSN:1006-6977/CN:61-1281/TN]

期数:
2016年02期
页码:
75-80
栏目:
出版日期:
2016-06-25

文章信息/Info

Title:
An Automated Lung Nodules based on Multidimensional Characteristics and the Support Vector Machine(SVM) Kernel Function Optimization
文章编号:
1672-6278 (2016)02-0075-06
作者:
刘银凤1△张俊杰2周涛2夏勇3吴翠颖2
1.宁夏医科大学附属心脑血管病医院 神经内科,宁夏 银川 750001;2.宁夏医科大学 理学院,宁夏 银川 750004;3.西北工业大学 计算机学院,陕西 西安 710100
Author(s):
LIU Yinfeng1ZHANG Junjie2ZHOU Tao2XIA Yong3WU Cuiying2
1.Neurology Department, Cardio-cerebrovascular Disease Hospital of Ningxia Medical University, Yinchuan 750001,China;2.School of Science, Ningxia Medical University, Yinchuan 750004;3.School of Computer Science, Northwestern Polytechnical University, Xi′an,Shaanxi 710100, China
关键词:
肺结节检测特征提取网格寻优支持向量机分类识别医学图像处理
Keywords:
Lung nodules detection Feature extraction Grid search Support Vector Machine(SVM) Classification and recognition Medical image processing
分类号:
R318
DOI:
-
文献标识码:
A
摘要:
为了解决特征级肺结节检测研究中的特征结构不合理和分类器性能低下两个问题,提出了一种多维特征表达与支持向量机(support vector machine, SVM)核函数优化相结合的自动化肺结节检测模型。首先提取多维特征数据量化感兴趣区域(region of interest, ROI),然后利用网格寻优算法优化SVM核函数,最后基于优化的SVM分类器识别结节区域和非结节区域。仿真实验结果表明,该模型耗时短、检测正确率高,具有一定的临床应用价值。
Abstract:
In order to solve the problems of unreasonable structure and poor classifier performance in the study of lung nodules detection based on feature level, an automatic model of lung nodules detection that multi-dimensional characteristic expression combined with SVM kernel function optimization was proposed. Firstly, multi-dimensional features were extracted to quantize ROI, and then grid search algorithm was used to optimize SVM kernel function, the SVM classifier was used to identify areas of nodules and others. The simulation results show that the proposed model uses less time and has high detection accuracy, which is valuable to clinical application.

参考文献/References

备注/Memo

备注/Memo:
收稿日期:2015-10-10 国家自然科学基金资助项目(81160183,61561040);宁夏自然科学基金资助项目(NZ14085);陕西省语音与图像信息处理重点实验室开放课题资助项目(SJ2013003) 通信作者Email:phoneixliu@163.com
更新日期/Last Update: 2016-06-30