DOCSLIB.ORG

Adaptive Filtering and Artificial Intelligence Methods on Fetal Ecg Extraction

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Adaptive Filtering and Artificial Intelligence Methods on Fetal Ecg Extraction Journal of Critical Reviews ISSN- 2394-5125 Vol 7, Issue 7, 2020 ADAPTIVE FILTERING AND ARTIFICIAL INTELLIGENCE METHODS ON FETAL ECG EXTRACTION Dr. M. Pradeepa1,Dr. S. Kumaraperumal2 1Assistant Professor (Sr.), School of Information Technology, Vellore Institute of Technology, Vellore, India. Email: [email protected] 2Sr. Assistant Professor , Xavier Institute of Management & Entrepreneurship, Bangalore, India. Email: [email protected] Corresponding Author Email ID [email protected] Received: 11.02.2020 Revised: 19.03.2020 Accepted: 23.04.2020 Abstract Above 30 percent of infant’s death occur due to heart problem like congenital heart disease during 2004 in United States of America. Every year, one in 125 infants is born with heart imperfection. To address these problems, early identification of cardiac anomalies and consistent monitoring of fetal heart can support Pediatric Cardiologist and Obstetrics to take necessary care on time to prescribe medicines and take precautionary measures during gestation period, delivery and/or after birth. Majority of cardiac abnormalities contain some symptoms in the cardiac electrical signal morphology. Electrocardiography gives more information in measuring cardiac signals compare to sonographic measurement. However, in non-invasive heartbeat recording by fetal Electrocardiogram (ECG) application Electrocardiography has its limitation due to low signal- noise ratio where impeding bio-signals are too stronger than fetal electrocardiogram signals. Various adaptive filtering and Artificial intelligence techniques are applied to solve this complex problem. The complex real world problems need a combination of knowledge, skills, and techniques from various sources as an intelligent system. That intelligent system should possess expertise of human, adjust itself to changing environment and learn to improve on its own. Keywords: Artificial Intelligent, Adaptive Filter, Adaptive Neuro-Fuzzy Inference System (ANFIS), Kalman filtering (KF). © 2019 by Advance Scientific Research. This is an open-access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/) DOI: http://dx.doi.org/10.31838/jcr.07.07.47 INTRODUCTION signals from more than one channels utilized to extract the fetal Fetus health condition is monitored by many methods where ECG components recorded at noisy environment. To extract Electrocardiography is one of the frequently used methods which desired signal many types of adaptive filters are used, in which shows the fetus heart’s electrical activities. some of them are presented below. Generally, an invasive or non-invasive method of recording of Time sequenced adaptive filtering has been recommended by Fetal ECG (FECG) is performed. In invasive method of recording, Ferrara & Widrow [4] for FECG enrichment. They identified the the electrode has to be placed on the scalp of the fetus to non-stationary fetal ECG signal having recurring statistical measure the ECG but the electrode has to be passed through characteristics. The Least Mean Square (LMS) adaptive filter can mother’s womb which creates difficulties to the mother [1] and able to follow up such fast changing non stationarities, hence an also possible only at the later stage of pregnancy period. The adaptive filter have to be designed with rapidly varying impulse non-invasive method of recording does not provide any trouble response to improve the performance of the extraction. The to the mother because the electrode has to be placed on mothers’ method uses many sets of hyper parameters to adapt for fast abdomen to measure the ECG of the fetus. changing impulse response. In order to adapt for fast changing impulse response, the method requires more abdominal signals There are several approaches proposed to record the fetal ECG and also timely identification of estimated fetal pulse. Apart from under non- invasive method which uses either a single lead or the above requirements the technique needs prior information of two leads or multiple leads. For a single lead method of fetal ECG positions. The time sequenced adaptive filtering recording, only one electrode is positioned on the mothers’ provides more accurate results compared to classical LMS abdomen, two lead systems uses two electrodes which have to be adaptive filter. The overall performance of the adaptive filter is positioned on the chest and abdomen and multiple lead systems increased when the number of channel input is increased. The require multiple electrodes to record the fetal ECG. main advantage of this approach is that the prior knowledge of There are several complications in non-invasive method of signals’ power spectrum is not required. But, the time sequenced recording fetal ECG, because the recording is not directly taken method need the estimation for the timely identification of the from the fetus which is measured on the abdomen, hence the pulse, to synchronize the filter regeneration and the fetal cardiac fetal ECG is to be extracted from signal contaminated by multiple cycles. They stated the future direction to enhance the results by sources of interferences. Apart from these sources of finding better method to locate the fetal pulse positions in order interferences the low signal level of fetal ECG [2] and the spectral to make this approach with recordings having lower SNR. overlapping of mother ECG and fetal ECG [3] makes the Kam & Cohen [5] identified a method to find the fetal ECG using extraction more critical. Infinite Impulse Response (IIR) filtering technique and Genetic Algorithm (GA). The hybrid IIR-GA approach on fetal ECG MATERIALS AND METHODS extraction, the adaptation rule is combined with GA, whenever Adaptive Filter based Methods the estimated gradient stuck with local extremum. Hybrid IIR-GA Generally, an adaptive filter has the ability of self-adjusting its provide best with simulation compared to FIR LMS based weight towards minimization of error. In this technique recorded method but with real data, the method fails to show the Journal of critical reviews 295 ADAPTIVE FILTERING AND ARTIFICIAL INTELLIGENCE METHODS ON FETAL ECG EXTRACTION significant difference between them. This may be because of the results than EKF. The performance evaluation was performed by body transfer function acts as a simple low pass filter so that a calculation accuracy, sensitivity, and positive predictive value. lower order FIR adaptive filter is sufficient, and the authors Single channel is used for the extraction which requires needs suggested further studies are required to analyze this few electronic components and suits for portable monitoring assumptions. system. Talha et al. [6] also presented similar approach of GA based Khamene & Negahdaripour [10] developed a method to extract Finite Impulse Response (FIR) filter for extraction where Genetic desired signal by wavelet transform using the modulus maxima algorithm is used as a optimizer for FIR filter and the results are which is in the wavelet domain and singularity detection, compared with the other approaches of adaptive filters like acquired from the abdomen signal. To differentiate the maternal wiener filter, Recursive Least Mean Square (RLMS) and NLMS and fetal ECG signals, abdominal signal’s modulus maxima filters. The NLMS approach provide better results in terms of locations are used. The authors projected two different reliability and speed of convergence but provide divergence approaches for implementation of the algorithm, in the first results when the adaptation is too large which have been approach, to carry out the classification minimum one thoracic overcome by the method of GA based FIR filter. GA with eight bits signal is used, but in the approach no thoracic signal is required. and ten iterations provide better quality compared to other A reconstruction algorithm is applied to extract the desired algorithms and an improvement may be provided by changing signal from the identified fetal modulus maxima. The developed the order of the filter. procedure varies from the traditional time domain approaches. In this method, the significant features of the signal provide high The adaptive filtering approach may be combined with other performance against other signal disturbances is studied. approaches to provide enhancement in extraction. Kholdi et al. [7] identified a GA based adaptive filter which uses LMS based Elloumi et al. [11] projected a fetal Electrocardiogram extraction adaptive filter for extracting fetal ECG, where the best filter through two stages Pitch Synchronous Wavelet Transform coefficients are calculated based on genetic algorithm which (PSWT). This implementation is based on a modeling idea which makes the adaptive filter response converge into global is capable of capturing the signal and its fluctuations by the use extremum. The random search nature of GA find the optimum of basic elements. In the two stage method, first the maternal filter coefficients even the structure of nonlinear transformation component estimation from a contaminated abdominal signal by is unknown or the structure may vary for different person. This a pitch synchronous decomposition is made. From the result of adaptive process does not need the knowledge in advance about first stage estimation, the next iteration is carried out to extract the signal and noise statistics, only assumes the signal is the required FECG signal. The pitch synchronous wavelet uncorrelated with noise. The input and output SNR are calculated decomposition permits to identify the
Load more

Recommended publications
  • Lossless Audio Coding Using Adaptive Linear Prediction
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by ScholarBank@NUS LOSSLESS AUDIO CODING USING ADAPTIVE LINEAR PREDICTION SU XIN RONG (B.Eng., SJTU, PRC) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF ENGINEERING DEPARTMENT OF ELECTRICAL AND COMPUTER ENGINEERING NATIONAL UNIVERSITY OF SINGAPORE 2005 ACKNOWLEDGEMENTS First of all, I would like to take this opportunity to express my deepest gratitude to my supervisor Dr. Huang Dong Yan from Institute for Infocomm Research for her continuous guidance and help, without which this thesis would not have been possible. I would also like to specially thank my supervisor Assistant Professor Nallanathan Arumugam from NUS for his continuous support and help. Finally, I would like to thank all the people who might help me during the project. ii TABLE OF CONTENTS ACKNOWLEDGEMENTS .............................................................................................. ii TABLE OF CONTENTS ................................................................................................. iii SUMMARY ........................................................................................................................vi LIST OF TABLES .......................................................................................................... viii LIST OF FIGURES ...........................................................................................................ix CHAPTER 1 INTRODUCTION.........................................................................................................
    [Show full text]
  • Total Harmonics Distortion Reduction Using Adaptive, Weiner, and Kalman Filters
    Western Michigan University ScholarWorks at WMU Master's Theses Graduate College 6-2016 Total Harmonics Distortion Reduction Using Adaptive, Weiner, and Kalman Filters Liqaa Alhafadhi Follow this and additional works at: https://scholarworks.wmich.edu/masters_theses Part of the Electrical and Computer Engineering Commons Recommended Citation Alhafadhi, Liqaa, "Total Harmonics Distortion Reduction Using Adaptive, Weiner, and Kalman Filters" (2016). Master's Theses. 737. https://scholarworks.wmich.edu/masters_theses/737 This Masters Thesis-Open Access is brought to you for free and open access by the Graduate College at ScholarWorks at WMU. It has been accepted for inclusion in Master's Theses by an authorized administrator of ScholarWorks at WMU. For more information, please contact [email protected]. TOTAL HARMONICS DISTORTION REDUCTION USING ADAPTIVE, WEINER, AND KALMAN FILTERS by Liqaa Alhafadhi A thesis submitted to the Graduate College in partial fulfillment of the requirements for the degree of Master of Science in Engineering (Electrical) Electrical and Computer Engineering Western Michigan University June 2016 Thesis Committee: Johnson Asumadu, Ph.D., Chair Massood Atashbar, Ph.D. Christopher Cho, Ph.D. TOTAL HARMONICS DISTORTION REDUCTION USING ADAPTIVE, WEINER, AND KALMAN FILTERS Liqaa Alhafadhi, M.S.E. Western Michigan University, 2016 Total harmonics distortion is one of the main problems in power systems due to its effects in generating undesirable issues in power quality. These effects include heating (in transformers, capacitors, motors, and generators), disoperation of electronic equipment, incorrect readings on meters, disoperation of protective relays, and communication interference. Besides these problems, harmonics affect the power quality in both transmission and distribution systems. Different techniques have been used to mitigate the effects of harmonics.
    [Show full text]
  • One-Dimensional Processing for Adaptive Image Restoration
    DOCUMENT OFFICE 36-412 JRESEARCH LABORATORY OF ELECTRONtCS *l MASSACHUSETTS INSTITUTE OF TECHNOLOGY ~~~~~~~~~IllLf llii One-Dimensional Processing for Adaptive Image Restoration Philip Chan tl 1; 4 eo'p ON// Technical Report 501 June 1984 Massachusetts Institute of Technology Research Laboratory of Electronics Cambridge, Massachusetts 02139 _ ___I __ _ __ __I_L *1 One-Dimensional Processing for Adaptive Image Restoration Philip Chan Technical Report 501 June 1984 Massachusetts Institute of Technology Research Laboratory of Electronics Cambridge, Massachusetts 02139 This work has been supported in part by the Advanced Research Projects Agency monitored by ONR under Contract N00014-81-K-0742 NR-049-506 and in part by the National Science Foundation under Grant ECS80-07102. _ __YLII··_I__IIL____yll_·ll-... -____. -- 1~ - - -- 1-~· 1 --- I I - ---·---- -- - _I __I __I UNCLASSIFTED SECUNITY CLASSIFICATION OF rIS PAGE REPORT OOCUMENTATION PAGE l& RPICORT SECURITY CLASSFPlCATION 1I RISTRlCTIVI MARKINGS 2& SICURITY C1ASSIFICATION AUTHORITY 3. OISTrIUUTIONIAVAILA6I1UTY OF RIPORT Approved for public release: distribution =6 O0CLAIf CATNoOWNGrAOlING sc.eout.a unlimited A. PRAFORMING ORGANIZATION IrEPoRTHMUSll(S) S. MONITORING ORGANIZATION REPORT NUM6EE(S) G. NAMIE OF PRFORlOMING ORGcANIZATION . OFICI( SYMKOL 7 NAM_ OF MONITORING ORGANIZATION Research Laboratory of Elec to mehe, Office of Naval Research Massachusetts Institute of Te:hnology Mathematical and Information Scien. Div. 6. A^OORSS (Ct. Saw emd ZIP Code) 7b. AOORESS =Ct. 31w .nJi ZIP Come 77 Massachusetts Avenue 800 North Quincy Street Cambridge, MA 02139 Arlington, Virginia 22217 &L NAM OF IINOIN(OWSPONSORING a OFICE SYMh t.oL L Ps1OCUtEMENT INSTRUMENT OENTIltCATION NUMllR ORGANIZATION (If WIawieal Advanced Research Projects .gency N00014-8 1-K-0742 Se.
    [Show full text]
  • Adaptive Filtering and Change Detection Fredrik Gustafsson Copyright © 2000 John Wiley & Sons, Ltd Isbns: 0-471-49287-6 (Hardback); 0-470-84161-3 (Electronic)
    Adaptive Filtering and Change Detection Fredrik Gustafsson Copyright © 2000 John Wiley & Sons, Ltd ISBNs: 0-471-49287-6 (Hardback); 0-470-84161-3 (Electronic) Adaptive Filtering and Change Detection Adaptive Filtering and Change Detection Fredrik Gustafsson Linkoping University, Linkoping, Sweden JOHN WILEY & SONS, LTD Chichester Weinheim NewYork Brisbane Singapore Toronto Copyright 0 2000 by John Wiley & Sons Ltd Baffins Lane, Chichester, West Sussex, PO 19lUD, England National 01243 779777 International (+44) 1243 779777 e-mail (for orders and customer service enquiries): [email protected] Visit our Home Page on http://www.wiley.co.uk or http://www.wiley.com All Rights Reserved.No part of this publication may be reproduced, storedin a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except under the termsof the Copyright Designs and PatentsAct 1988 or under the termsof a licence issuedby the Copyright Licensing Agency,90 Tottenham Court Road, London, W1P 9HE,UK, without the permissionin writing of the Publisher, with the exceptionof any material supplied specifically for the purposeof being entered and executed on a computer system, for exclusive use by the ofpurchaser the publication. Neither the authors nor John Wiley& Sons Ltd accept any responsibility or liability for loss or damage occasioned to any person or property through using the material, instructions, methods or ideas contained herein, or acting or refraining from actingas a result of such use. The authors and Publisher expressly disclaim all implied warranties, including merchantabilityof fitness for any particular purpose. There will be no duty on the authorsof Publisher to correct any errorsor defects in the software.
    [Show full text]
  • Image Restoration with Two-Dimensional Adaptive Filter
    Image Denoising with Two-Dimensional Adaptive Filter Algorithms M. Shams Esfand Abadi* and S. Nikbakht* Abstract: Two-dimensional (2D) adaptive filtering is a technique that can be applied to many image and signal processing applications. This paper extends the one-dimensional adaptive filter algorithms to 2D structure and the novel 2D adaptive filters are established. Based on this extension, the 2D variable step-size normalized least mean squares (2D-VSS- NLMS), the 2D-VSS affine projection algorithms (2D-VSS-APA), the 2D set-membership NLMS (2D-SM-NLMS), the 2D-SM-APA, the 2D selective partial update NLMS (2D- SPU-NLMS), and the 2D-SPU-APA are presented. In 2D-VSS adaptive filters, the step- size changes during the adaptation which leads to improve the performance of the algorithms. In 2D-SM adaptive filter algorithms, the filter coefficients are not updated at each iteration. Therefore, the computational complexity is reduced. In 2D-SPU adaptive algorithms, the filter coefficients are partially updated which reduce the computational complexity. We demonstrate the good performance of the proposed algorithms thorough several simulation results in 2D adaptive noise cancellation (2D-ANC) for image denoising. The results are compared with the classical 2D adaptive filters such as 2D-LMS, 2D-NLMS, and 2D-APA. Keywords: Two Dimensional, Adaptive Filter, Noise Cancellation, Selective Partial Update, Variable Step-Size, Set-Membership. 1 Introduction` one direction, was proposed in [7]. Also, the Adaptive filter algorithms have numerous applications development of a 2D adaptive filter using the block in electrical engineering [1], [2] and [3]. Two- diagonal LMS method was presented in [8].
    [Show full text]
  • Determination of Optimum Step Size of LMS Algorithm in De-Noising of Image
    Determination of Optimum Step Size of LMS Algorithm in De-noising of Image This Major project in partial fulfillment for the award of the degree of Bachelor of Science In Information and Communication Engineering Department of ECE, Faculty of Science and Engineering East West University, Dhaka Bangladesh Developed By Sharmin Akter Mim #ID: 2013-2-50-002 Nusrat Jahan #ID: 2012-3-50-001 Under the supervision of Dr. Mohamed Ruhul Amin Professor Dept. of Electronics and Communication Engineering East West University & Dr. Md. Imdadul Islam Professor Dept. of Computer Science and Engineering Jahangirnagar University Dept. of ECE @ EWU Page 1 DECLARATION We hereby declare that we carried out the work reported in this project in the Department ofElectronics and Communications Engineering, East West University, under the supervision ofDr. Mohamed Ruhul Amin and Dr. Md. Imdadul Islam. We solemnly declare that to the best of our knowledge, no part of this report has been submitted elsewhere for award of a degree. All sources of knowledge usedhave been duly acknowledged. Signature: ----------------------------------------- ------------------------------------------ SharminAkterMim Nusrat Jahan ID: 2013-2-50-002 ID: 2012-3-50-001 ------------------------------------- Supervisor Dr. Mohamed Ruhul Amin Professor, Dept. of Electronics and Communication Engineering East West University & ------------------------------------- Supervisor Dr. Md. Imdadul Islam Professor, Dept. of Computer Science and Engineering Jahangirnagar University Dept. of ECE @ EWU Page 2 CERTIFICATE This is to certify that the major project entitled “Determination of Optimum Step Size of LMS Algorithm in De-noising of Image” being submitted by Sharmin Akter Mim and Nusrat Jahan of Electronics and Communications Engineering Department, East West University, Dhaka in partial fulfillment for the award of the degree of Bachelor of Science in Information and Communication Engineering, is a record of major project carried out by them.
    [Show full text]
  • Adaptive Digital Filters Branko Kovacˇevic´ • Zoran Banjac Milan Milosavljevic´
    Adaptive Digital Filters Branko Kovacˇevic´ • Zoran Banjac Milan Milosavljevic´ Adaptive Digital Filters Academic Mind University of Belgrade - School of Electrical Engineering University Singidunum, Belgrade Springer-Verlag Berlin Heidelberg Branko Kovacˇevic´ Milan Milosavljevic´ University of Belgrade University of Belgrade and Singidunum Belgrade University Serbia Belgrade Serbia Zoran Banjac School of Electrical and Computing Engineering of Applied Studies Belgrade Serbia ISBN 978-3-642-33560-0 ISBN 978-3-642-33561-7 (eBook) DOI 10.1007/978-3-642-33561-7 Springer Heidelberg New York Dordrecht London Library of Congress Control Number: 2013935402 ISBN of Academic Mind : 978-86-7466-434-6 Ó Academic Mind Belgrade and Springer-Verlag Berlin Heidelberg 2013 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer and Academic Mind. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center.
    [Show full text]
  • Implementation of Adaptive Filtering Algorithms for Noise Cancellation
    FACULTY OF ENGINEERING AND SUSTAINABLE DEVELOPMENT Department of Electronics, Mathematics and Natural Sciences Implementation of adaptive filtering algorithms for noise cancellation Tanja Lampl 2020 Student thesis, Advanced level (Master degree), 15 HE Electronics Master Programme in Electronics/Automation (online) Supervisor: José Chilo (PhD) Examiner: Niclas Björsell (PhD) i Abstract This paper deals with the implementation and performance evaluation of adaptive filtering algorithms for noise cancellation without reference signal. Noise cancellation is a technique of estimating a desired signal from a noise- corrupted observation. If the signal and noise characteristics are unknown or change continuously over time, the need of adaptive filter arises. In contrast to the conventional digital filter design techniques, adaptive filters do not have constant filter parameters, they have the capability to continuously adjust their coefficients to their operating environment. To design an adaptive filter, that produces an optimum estimate of the desired signal from the noisy environment, different adaptive filtering algorithms are implemented and compared to each other. The Least Mean Square LMS, the Normalized Least Mean Square NLMS and the Recursive Least Square RLS algorithm are investigated. Three performance criteria are used in the study of these algorithms: the rate of convergence, the error performance and the signal-to- noise ratio SNR. The implementation results show that the adaptive noise cancellation application benefits more from
    [Show full text]
  • Analysis of Adaptive Filter Algorithms with an Application to Harmonic
    Analysis of Adaptive Filter Algorithms with an Application to Harmonic Noise Cancellation in Distribution Power Line Communications by Jin-Der Wang Center for Communications and Signal Processing Department of Electrical and Computer Engineering North Carolina State University August 1985 CCSP-TR-85/15 ABSTRACT WANG, JIN-DER. Analysis of Adaptive filter Algorithms with an Application to Harmonic Noise Cancellation in Distribution Power Line Communications. (Under the direction of Dr. Henry Joel Trussell.) Some well-known adaptive algorithms such as the least mean squares (LMS) and the recursive least squares (RLS) algorithms are reviewed, and the properties of some newly developed fast RLS algorithms such as the fast Kalman (FK), the fast a priori estimation sequential technique (FAEST), and the fast transversal filter (FTF) algorithms are investigated. These adaptive algorithms are applied to the problem of harmonic noise cancellation in distribution power line communica­ tions. The fast RLS algorithms were derived independently, from different approaches, and no connection had previously been made. They are rederived with a unified approach, and their mathematical equivalence is shown by examining certain algorithmic quantities and initial conditions. Based on this work, an improved rescue variable is then proposed which can detect the tendency of algo­ rithm divergence earlier than previously proposed rescue variables. Adaptive filters are applied to harmonic noise cancellation m distribution power line communications. It is shown that most filter taps can be constrained to zero to reduce computational complexity. The constrained LMS and RLS algo­ rithms are derived for a general constraint problem, and the constrained FK algorithm is derived for harmonic noise cancellation.
    [Show full text]