Age-related alterations in the sign series entropy of short-term pulse rate variability

详细信息    查看全文

• 作者：Yongxin Chou^a ; ^b ; ^{lutchouyx@163.com}Author Vitae ; Aihua Zhang^a ; ^c ; ^{lutzhangah@163.com}Author Vitae ; Bin Yang^a ; ^c ; ^{yangbin0@yeah.net}Author Vitae
• 关键词：Short-term pulse rate variability ; Sign series entropy ; Sliding window iterative sign series entropy analysis ; Probabilistic neural network ; Age classification
• 刊名：Neurocomputing
• 出版年：2017
• 出版时间：8 March 2017
• 年：2017
• 卷：228
• 期：Complete
• 页码：213-219
• 全文大小：968 K
• 卷排序：228

文摘

Pulse rate variability (PRV) has been certified as a substitute for heart rate variability (HRV) to diagnose and predict some diseases. Sign series entropy analysis (SSEA), a kind of nonlinear method, has been used to analyze HRV signal effectively. However, the time consumption of SSEA is too long, and it is unknown whether SSEA is suitable for analyzing short-term PRV signals. Therefore, an improved SSEA method named sliding window iterative SSEA (SWISSEA) is proposed to analyze short-term PRV signal and derive some age-related alterations. Moreover, a radial basis probabilistic neural network (RBPNN) based algorithm is proposed to classify subjects according to their ages. Continuous non-invasive blood pressure signals from the MIT-BIH database are chosen to generate short-term PRV signals as the experimental data, and their time domain and frequency domain parameters are extracted and selected for classifying. The experimental results show that the pulse beats are more uniform as the increase of the age, the sign series entropy (SSE) increases with aging and has a significant difference between young and old subjects even if the PRV signal is corrupted by heavy noises, the approach based on RBPNN can accurately classify subjects according their ages. In addition, the SWISSEA reduces the time consumption of SSEA, is more suitable for analyzing short-term PRV signals in real time, and has a potential in portable medical devices.