Accurate and reliable 3-lead to 12-lead ECG reconstruction methodology for remote health monitoring applications

Abstract Remote areas, generally, encounter scarcity of cardiologists and state-of-the-art facilities for the treatment of cardiovascular diseases (CVD). However, aforementioned constraints can be addressed using recent advancements in wireless technology and its omnipresence. Clinical CVD diagnosis is, generally, carried out using either of Standard 12-Lead (S12) or Mason–Likar 12-Lead (ML12) system. These systems consist of 8 independent leads/signals which restrict their usage in telemonitoring applications viz. personalized remote health monitoring, home monitoring etc., due to high bandwidth and storage requirements, data transmission time and low compression ratio (CR) from signal compression techniques. Moreover, the reduced lead (RL) systems with 2–3 leads, generally, employed in telemonitoring applications might not be sufficient for diagnosis. We form a Reduced 3-Lead system from S12 and ML12 systems at the transmission end which reduces the number of signals to 3 and then reconstruct the S12 and ML12 systems at the receiver end using our proposed personalized reconstruction methodology, thus allaying aforementioned limitations. I, II and V 2 form the basis leads and the precordial leads form the target leads. Least square fit and heart vector projection theory have been used to obtain personalized transformation coefficients. Accuracy of reconstruction has been evaluated on PhysioNet PTBDB and INCARTDB, after wavelet based preprocessing, using R 2 statistics, correlation ( r x ) and regression ( b x ) coefficients. Re-usability of personalized coefficients has also been investigated in this paper. Mean R 2 values obtained from the reconstruction of target leads are 91.87% (PTBDB) 83.75% (INCARTDB). R3L system reduces the number of leads/signals from 8 to 3 and as the results indicate the possibility of re-using the transformation coefficients, the number of electrodes can be reduced from 10 to 5, thereby, increasing the comfort of patients and caregivers.

[1]  Jeffrey M. Hausdorff,et al.  Physionet: Components of a New Research Resource for Complex Physiologic Signals". Circu-lation Vol , 2000 .

[2]  Gian Franco Gensini,et al.  A Clinician’s View of Next-Generation Remote Healthcare System , 2014 .

[3]  Amit Acharyya,et al.  Development of an Automated Updated Selvester QRS Scoring System Using SWT-Based QRS Fractionation Detection and Classification , 2014, IEEE Journal of Biomedical and Health Informatics.

[4]  Jun Lin,et al.  The Optimal De-noising Algorithm for ECG Using Stationary Wavelet Transform , 2009, 2009 WRI World Congress on Computer Science and Information Engineering.

[5]  G E Dower,et al.  A lead synthesizer for the Frank system to simulate the standard 12-lead electrocardiogram. , 1968, Journal of electrocardiology.

[6]  Koushik Maharatna,et al.  Robust and accurate personalised reconstruction of standard 12-lead system from Frank vectorcardiographic system , 2016, Comput. methods Biomech. Biomed. Eng. Imaging Vis..

[7]  Amit Acharyya,et al.  Reduced lead system selection methodology for reliable standard 12-lead reconstruction targeting personalised remote health monitoring applications , 2014, Comput. methods Biomech. Biomed. Eng. Imaging Vis..

[8]  Stefan P Nelwan,et al.  Reconstruction of the 12-lead electrocardiogram from reduced lead sets. , 2004, Journal of electrocardiology.

[9]  R Hoekema,et al.  On selecting a body surface mapping procedure. , 1999, Journal of electrocardiology.

[10]  Koushik Maharatna,et al.  An automated algorithm for online detection of fragmented QRS and identification of its various morphologies , 2013, Journal of The Royal Society Interface.

[11]  James W Warren,et al.  On designing and testing transformations for derivation of standard 12-lead/18-lead electrocardiograms and vectorcardiograms from reduced sets of predictor leads. , 2008, Journal of electrocardiology.

[12]  J. Estes,et al.  Abdominal Aortic Aneurysm: A Study of One Hundred and Two Cases , 1950 .

[13]  M. Schalij,et al.  Reconstruction of standard 12-lead electrocardiograms from 12-lead electrocardiograms recorded with the Mason-Likar electrode configuration. , 2008, Journal of electrocardiology.

[14]  Richard E Gregg,et al.  Technical challenges and future directions in lead reconstruction for reduced-lead systems. , 2008, Journal of electrocardiology.

[15]  Amit Acharyya,et al.  A Low-Complexity ECG Feature Extraction Algorithm for Mobile Healthcare Applications , 2013, IEEE Journal of Biomedical and Health Informatics.

[16]  Hui Yang,et al.  Linear affine transformations between 3-lead (Frank XYZ leads) vectorcardiogram and 12-lead electrocardiogram signals. , 2009, Journal of electrocardiology.

[17]  Ernest Frank,et al.  General Theory of Heart‐Vector Projection , 1954 .

[18]  Chris D Nugent,et al.  Synthesising the 12-lead electrocardiogram: Trends and challenges. , 2007, European journal of internal medicine.

[19]  Samarendra Dandapat,et al.  Multichannel ECG Data Compression Based on Multiscale Principal Component Analysis , 2012, IEEE Transactions on Information Technology in Biomedicine.

[20]  G E Dower,et al.  Deriving the 12-lead electrocardiogram from four (EASI) electrodes. , 1988, Journal of electrocardiology.

[21]  E. Helfenbein,et al.  Limitations on the Re-use of patient specific coefficients for 12-lead ECG reconstruction , 2008, 2008 Computers in Cardiology.

[22]  J A Kors,et al.  Minimal lead sets for reconstruction of 12-lead electrocardiograms. , 2000, Journal of electrocardiology.

[23]  Sophia H Zhou,et al.  Where do derived precordial leads fail? , 2008, Journal of electrocardiology.

[24]  David S. Rosenbaum,et al.  Keeping a beat on the heart , 2004, IEEE Pervasive Computing.

[25]  Dewar D Finlay,et al.  Standardization of reduced and optimal lead sets for continuous electrocardiogram monitoring: where do we stand? , 2008, Journal of electrocardiology.

[26]  Pachamuthu Rajalakshmi,et al.  Accurate and reliable 3-lead to 12-lead ECG reconstruction methodology for remote health monitoring applications , 2013, 2013 IEEE 15th International Conference on e-Health Networking, Applications and Services (Healthcom 2013).

[27]  Anestis Antoniadis,et al.  Wavelet Estimators in Nonparametric Regression: A Comparative Simulation Study , 2001 .

[28]  Stefan P Nelwan,et al.  Assessment of derived 12-lead electrocardiograms using general and patient-specific reconstruction strategies at rest and during transient myocardial ischemia. , 2004, The American journal of cardiology.

[29]  J. A. Scherer,et al.  Synthesis of the 12 lead electrocardiogram from a 3 lead semi-orthogonal subset using patient-specific linear transformation arrays , 1988, Proceedings. Computers in Cardiology 1988.

[30]  K. Maharatna,et al.  A Time-Domain Morphology and Gradient based algorithm for ECG feature extraction , 2012, 2012 IEEE International Conference on Industrial Technology.

[31]  Amit Acharyya,et al.  Methodology for automated detection of fragmentation in QRS complex of Standard 12-lead ECG , 2013, 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC).

[32]  Ralf Bousseljot,et al.  Nutzung der EKG-Signaldatenbank CARDIODAT der PTB über das Internet , 2009 .

[33]  Ch. L. Levkov,et al.  Orthogonal electrocardiogram derived from the limb and chest electrodes of the conventional 12-lead system , 1987, Medical and Biological Engineering and Computing.

[34]  Donghui Zhang,et al.  Wavelet Approach for ECG Baseline Wander Correction and Noise Reduction , 2005, 2005 IEEE Engineering in Medicine and Biology 27th Annual Conference.