Please use this identifier to cite or link to this item:
Title: Nonlinear methods and analysis of heart rate variability
Researcher: Santhi C
Guide(s): Kumaravel, N.
Keywords: Nonlinear methods, Heart Rate Variability, Adaptive Threshold Rank Order Filter, Expectation Maximization (EM) Algorithm
Upload Date: 28-Nov-2013
University: Anna University
Completed Date: 
Abstract: Variation in beat-to-beat interval of heart is known as heart rate variability (HRV). HRV offers a non-invasive method of evaluating input into cardiac rhythm. The principal objective of this research work is to discriminate healthy and pathological subjects HRV using adaptive nonlinear methods. Proposed works in this thesis are: Adaptive Threshold Rank Order Filter (ATROF) to remove the HRV artifacts; an adaptive de-trending technique to process the nonstationary trend and its effect on very low frequency variations of HRV; Latency analysis of HRV in Different Age and Pathological Conditions. The analysis of Heart Rate Variability (HRV) demands specific capabilities which are not provided either by parametric or nonparametric conventional estimation methods. The statistical analysis of heart rate variability shows a significant departure from normality as reflected by the excess kurtosis, skew ness and rejection of normality in hypothetical tests. The heart rate variability is decomposed into three different variances (and#945;, Ø, Q) and estimated through a nonlinear non-Gaussian state space model and parameter estimation method. Variance and#945; represents the long run average; Variance Ø represents the fluctuation strength and variance Q represents the innovation that was not available when previous measurements were made. The parameters of the model are estimated using Particle filtering, Particle smoothing and Expectation Maximization (EM) Algorithm. In the simulation study it has been shown that the estimated parameters are close to the true parameters of the model, which shows the suitability of parameter estimation method for the described model. In the data analysis, it is found that the fluctuation strength and variance of innovation are highly correlated to mean value of fluctuations. The proposed nonlinear techniques are adaptive to the complexity of HRV signal and do not assume any functional forms or scales prior to analysis. newline newline newline
Pagination: xxvii,203
Appears in Departments:Faculty of Information and Communication Engineering

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File49.31 kBAdobe PDFView/Open
02_certificates.pdf443.9 kBAdobe PDFView/Open
03_abstract.pdf25.61 kBAdobe PDFView/Open
04_acknowledgement.pdf13.31 kBAdobe PDFView/Open
05_contents.pdf69.12 kBAdobe PDFView/Open
06_chapter 1.pdf285.14 kBAdobe PDFView/Open
07_chapter 2.pdf2.48 MBAdobe PDFView/Open
08_chapter 3.pdf211.84 kBAdobe PDFView/Open
09_chapter 4.pdf194.42 kBAdobe PDFView/Open
10_chapter 5.pdf360.52 kBAdobe PDFView/Open
11_chapter 6.pdf233.26 kBAdobe PDFView/Open
12_chapter 7.pdf68.27 kBAdobe PDFView/Open
13_appendices 1 to 3.pdf238.57 kBAdobe PDFView/Open
14_references.pdf44.64 kBAdobe PDFView/Open
15_publications.pdf17.48 kBAdobe PDFView/Open
16_vitae.pdf10.39 kBAdobe PDFView/Open

Items in Shodhganga are protected by copyright, with all rights reserved, unless otherwise indicated.