基于句法模式识别的脉搏波特征信息研究  

Application of Syntactic Pattern Recognition in Research on Pulse Wave's Characteristic Information

作　　者：黄祥钧[1] 邢武 李锋[1] 

机构地区：[1]中国科学院合肥智能机械研究所,安徽合肥230031

出　　处：《中国医疗器械杂志》2005年第5期325-327,共3页Chinese Journal of Medical Instrumentation

摘　　要：提出一种基于句法模式识别的脉搏波特征信息检测方法。该方法对于无创伤检测到的脉搏波信号,经过分段线形化基元提取,得到输入模式,然后使用Earley算法构造识别器对脉搏波波形进行模式判别,并在模式信息的基础上判断识别脉搏波特征信息,从而为心血管功能诊断提供决策依据。实验表明,该方法抗干扰能力强、准确度高,实现了波形识别的智能化,为脉诊系统的智能化提供了一条先进而实用的手段。In this paper, a syntactic pattern recognition method of the detection of pulse wave＇s characteristic information is introduced. According to it, the pressure signal of pulse is measured by non-invasive diagnostic method. Then, the elementary abstracting system transfers it into an input pattern. The recognizer, constructed by Earley algorithm, recognizes the pulse wave＇s pattern. On the base of it, the pulse wave＇s characteristic information can be got. Thus, the expert system will make a diagnosis with the information. The experiment result shows that, the new method is of anti-interference, preciseness and intelligence, and it provides an advanced, practical and intelligent means for the pulse diagnostic system.

关 键 词：脉搏波 句法模式识别 基元 识别器 

分 类 号：R318.04[医药卫生—生物医学工程]