| 神经网络在骨密度超声波检测中的应用 |
| |
| 引用本文： | 胡海燕,张荣标,柏受军. 神经网络在骨密度超声波检测中的应用[J]. 北京生物医学工程, 2007, 26(2): 187-190 |
| |
| 作者姓名： | 胡海燕 张荣标 柏受军 |
| |
| 作者单位： | 江苏大学电气信息工程学院,江苏镇江,212013;江苏大学电气信息工程学院,江苏镇江,212013;江苏大学电气信息工程学院,江苏镇江,212013 |
| |
| 摘 要： | 定量超声骨密度检测方法是直接利用超声波信号的参数变化评价骨密度特性,它存在参数映射关系模糊、检测结果可信度低的问题.提出利用人工神经网络方法,建立不同年龄段正常女性跟骨超声波信号的参数变化与标准骨密度参考数据库之间的网络模型,采用Bayesian正则化算法训练网络.结果表明,建立的神经网络模型反映了不同年龄段超声波检测参数变化与标准骨密度参考值之间良好的映射能力,提高了定量超声法检测骨密度的可信性.
|
| 关 键 词： | 超声波声速 衰减 骨密度 BP神经网络 |
| 文章编号： | 1002-3208（2007）02-0187-04 |
| 收稿时间： | 2006-03-14 |
| 修稿时间： | 2006-04-26 |
Application of the neural network in the detection of bone mineral density using ultrasonic |
| |
| HU Haiyan,ZHANG Rongbiao,BAI Shoujun. Application of the neural network in the detection of bone mineral density using ultrasonic[J]. Beijing Biomedical Engineering, 2007, 26(2): 187-190 |
| |
| Authors: | HU Haiyan ZHANG Rongbiao BAI Shoujun |
| |
| Affiliation: | Shoujun Institute of Electrical and information Engineering, Jiangsu University, Zhenjiang, Jiangsu Province 212013 |
| |
| Abstract: | Parameters of ultrasound are directly used to evaluate the characteristic of bone mineral density using Quantitative ultrasound method.This method exsited the problems of dark mapped relationship and low reliability.The neural network method is proposed in this paper.The network model is estabilished on the relationship of the ultrasound parameters and the normal database of bone mineral density for different aged women.The network is trained with the Bayesian method.The results showed ,with the neural network model,the mapped relationship of the ultrasound parameters and the normal database of bone mineral density for the different ages could be reflected and the creditability of detecting bone mineral density using ultrasound is improved. |
| |
| Keywords: | ultrasound velocity ultrasound attenuation bone mineral density back-proragation neural network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
