摘要
目的利用31P磁共振波谱分析数据,区别肝细胞肝癌、肝硬化和正常的肝组织。方法从33例包括肝细胞肝癌、肝硬化和正常肝的志愿者中选择66个有效体素样本,利用1.5T超导MR扫描仪采集31P MRS数据,通过BP神经网络和SVM模型的实验来区别肝细胞肝癌、肝硬化和正常肝脏组织。结果有限的样本实现了良好的分类性能,反向传播神经网络(BP)和支持向量机(SVM)模型可以提高31P MRS识别率,识别率可达92.31%。结论基于BP和SVM的31P MRS数据分析,对于活体肝细胞肝癌的诊断提供了一种可选择的有价值的技术。
Objective To distinguish hepatocellular carcinoma, liver cirrhosis and normal liver tissues using ^31P MRS. Methods Using a 1.5T superconducting MR scanner, ^31P MRS data of 66 effective voxel samples were collected from 33 cases of hepatocellular carcinoma, liver cirrhosis and normal liver volunteers. The BP neural network and SVM model experiments were used to differentiate hepatocellular carcinoma, liver cirrhosis and normal liver tissues. Results Linfited samples achieved a good classification performance. Back-propagation neural network （BP） and Support Vector Machine （SVM） models can improve identification rates of ^31P MRS to 92.31%. Conclusion ^31P MRS data analysis based on BP and SVM for in vivo diagnosis of hepatoeellular carcinoma provides a valuable optional technology.
出处
《山东大学学报（医学版）》
CAS
北大核心
2009年第6期42-46,共5页
Journal of Shandong University:Health Sciences
基金
山东省自然科学基金资助项目(Y2006C96)
关键词
癌
肝细胞
肝硬化
磁共振波谱学
支持向量机
反向传播神经网络
Carcinoma, hepatocellular
liver cirrhosis
magnetic resonance spectroscopy
Support vector machine
Back-propagation neural network
