摘要： 提出一种基于可变宽度模板的多模板匹配的室性早搏判别算法。该方法通过样本数据创建初始模板库，根据不同的RR间期确定模板宽度，使用基于模板队列匹配和模板库匹配的二级匹配策略，利用相关系数和RR间期比相结合的算法检测室早，模板库可自动扩充。经MIT-BIH心律失常数据库测试，该算法达到了98.66%的灵敏度(正确率)，平均错误率为1.11%。
关键词:
多模板匹配,
可变宽度模板,
RR间期,
室性早搏
Abstract: An algorithm based on multiple variable-width template matching for Premature Ventricular Contraction(PVC) recognition is presented. A template database is created by means of training samples, in which the width of each template is based on its RR interval. A two-stage strategy is proposed, which is based on template queue matching and template database matching, combining correlation coefficient with ratio of RR intervals. And the new template can be added to the database automatically. The performance of this algorithm is evaluated on the MIT-BIH arrhythmia database and it shows that the sensitivity for PVC reaches 98.66%, and AER is 1.11%.
Key words:
multiple template matching,
variable-width template,
RR interval,
Premature Ventricular Contraction(PVC)
中图分类号:
杨波, 张跃. 基于多模板匹配的室性早搏判别算法[J]. 计算机工程, 2010, 36(16): 291-292.
YANG Bei, ZHANG Ti. PVC Recognition Algorithm Based on Multiple Template Matching[J]. Computer Engineering, 2010, 36(16): 291-292.