Predicting the QRS complex and detecting small changes using principal component analysis
-
Antoun Khawaja
Abstract
In this paper, a new method for QRS complex analysis and estimation based on principal component analysis (PCA) and polynomial fitting techniques is presented. Multi-channel ECG signals were recorded and QRS complexes were obtained from every channel and aligned perfectly in matrices. For every channel, the covariance matrix was calculated from the QRS complex data matrix of many heartbeats. Then the corresponding eigenvectors and eigenvalues were calculated and reconstruction parameter vectors were computed by expansion of every beat in terms of the principal eigenvectors. These parameter vectors show short-term fluctuations that have to be discriminated from abrupt changes or long-term trends that might indicate diseases. For this purpose, first-order poly-fit methods were applied to the elements of the reconstruction parameter vectors. In healthy volunteers, subsequent QRS complexes were estimated by calculating the corresponding reconstruction parameter vectors derived from these functions. The similarity, absolute error and RMS error between the original and predicted QRS complexes were measured. Based on this work, thresholds can be defined for changes in the parameter vectors that indicate diseases.
References
[1] Ambroggi L. T wave complexity: diagnostic and prognostic significance. Int J Bioelectromagn2003; 5: 297–299.Search in Google Scholar
[2] Khawaja A, Sanyal S, Dössel O. A wavelet-based technique for baseline wander correction in ECG and multi-channel ECG. In: Proceedings of NBC'05, Vol. 9. Sweden: IFMBE 2005: 291–292.Search in Google Scholar
[3] Khawaja A, Sanyal S, Dössel O. A wavelet-based multi-channel ECG delineator. In: Proceedings of EMBEC'05, Vol. 11. IFMBE 2005: CD-ROM, unpaginated.Search in Google Scholar
[4] Jolliffe IT. Principal component analysis. New York: Springer 2002: 10–76.Search in Google Scholar
[5] Principal component analysis. www.ucl.ac.uk/oncology/MicroCore/HTML_resource/PCA_1.htm (accessed in Oct-ober 2005).Search in Google Scholar
[6] Principal component analysis. http://www.fon.hum.uva.nl/praat/manual/Principal_component_analysis.html (ac-cessed in October 2005).Search in Google Scholar
[7] Lux RL, Evans AK, Burgess MJ, Wyatt RF, Abildskov JA. Redundancy reduction for improved display and analysis of body surface potential maps. I. Spatial compression. Circ Res1981; 49: 186–196.10.1161/01.RES.49.1.186Search in Google Scholar
[8] Emdin M, Taddei A, Varanini M, et al. Compact representation of autonomic stimulation on cardiorespiratory signals by principal component analysis. IEEE 1993 Computers in Cardiology Conference. Piscataway, NJ: IEEE Press 1993: 157–160.Search in Google Scholar
[9] Presedo J, Fernandez EA, Vila J, Barro S. Cycles of ECG parameter evolution during ischemic episodes. IEEE 1996 Computers in Cardiology Conference. Piscataway, NJ: IEEE Press 1996: 489–492.Search in Google Scholar
[10] Sandham WA, Thomson DC, Hamilton DJ. ECG compression using artificial neural networks. In: Proceedings of the IEEE 17th International Conference of the Engineering in Medicine and Biology Society.Piscataway, NJ: IEEE Press 1995: 193–194.Search in Google Scholar
[11] Nagasaka Y, Iwata A. Performance evaluation of BP and PCA neural networks for ECG data compression. In: Proceedings of the 1993 International Joint Conference on Neural Networks. 2003: 1003–1006.Search in Google Scholar
[12] Teodorescu HN, Bonciu C. Feedforward neural filter with learning in features space: preliminary results. In: Proceedings of the International Symposium on Neuro-Fuzzy AT'96. Lausanne 1996: 17–24.Search in Google Scholar
[13] Costa EV, Moraes J. QRS feature discrimination capability: quantitative and qualitative analysis. IEEE Comput Cardiol2000; 27: 399–402.Search in Google Scholar
[14] Stamkopoulos T, Maglaveras N. ECG analysis using nonlinear PCA neural networks for ischemia detection. IEEE Trans Signal Process1998; 46: 3058–3067.10.1109/78.726818Search in Google Scholar
[15] Vargas F, Lettnin D, de Castro M, Macarthy M. Electrocardiogram pattern recognition by means of MLP network and PCA: a case study on equal amount of input signal types. In: Proceedings of SBRN'02. 2002: 200–205.Search in Google Scholar
[16] Panerai R, Ferreira L, Brum O. Principal component analysis of multiple noninvasive blood flow derived signals. IEEE Trans Biomed Eng1988; 35: 533–538.10.1109/10.4582Search in Google Scholar PubMed
[17] Newandee D, Reisman S, Bartels MN, Meersman RE. COPD severity classification using principal component and cluster analysis on HRV parameters. In: Proceedings of the 29th Bioengineering Conference. 2003: 134–135.Search in Google Scholar
[18] Ye WY, Li G, Lin L, Yu QL. ECG analysis based on PCA and SOM. In: Proceedings of the 2003 IEEE International Conference on Neural Networks and Signal Processing. Piscataway, NJ: IEEE Press 2003: 37–40.Search in Google Scholar
[19] Okin P, Devereux R, Fabsitz R, Lee E, Galloway J, Howard B. Principal component analysis of the T wave and prediction of cardiovascular mortality in American Indians. Circulation2002; 105: 714–719.10.1161/hc0602.103585Search in Google Scholar PubMed
[20] Severi S, Vecchietti S, Cavalcanti S, Santoro A, Bie J. Analysis of T wave complexity during arrhythmic phases caused by hemodialysis. IEEE Comput Cardiol2001; 28: 637–640.Search in Google Scholar
[21] Khawaja A, Dössel O. A PCA-based technique for QRS complex estimation. IEEE Comput Cardiol2005; 32: 747–750.10.1109/CIC.2005.1588212Search in Google Scholar
[22] Köhler BU, Hennig C, Orglmeister R. The principles of software QRS detection. IEEE Eng Med Biol2002; 21: 42–49.10.1109/51.993193Search in Google Scholar PubMed
[23] Smith LI. A tutorial on principal component analysis. http://www.cs.otago.ac.nz/cosc453/student_tutorials/principal_components.pdf (accessed in October 2005).Search in Google Scholar
[24] Curve fitting toolbox: user's guide. Natick, MA: The MathWorks Inc 2005: 13–59.Search in Google Scholar
[25] Statistics toolbox: user's guide. Natick, MA: The MathWorks Inc 2005: 149–165.Search in Google Scholar
©2007 by Walter de Gruyter Berlin New York
Articles in the same Issue
- Ralph Mueller and Herbert Witte join the Associate Editor team of Biomedizinische Technik/Biomedical Engineering
- Technological innovations in information engineering demand sustained updating and upgrading in biosignal processing applications: a continual renaissance
- Predicting initiation and termination of atrial fibrillation from the ECG
- Predicting the QRS complex and detecting small changes using principal component analysis
- The role of independent component analysis in the signal processing of ECG recordings
- Implantable cardioverter defibrillator algorithms: status review in terms of computational cost
- Assessment of dynamic changes in cerebral autoregulation
- Corrected body surface potential mapping
- Autonomic cardiac control in animal models of cardiovascular diseases. I. Methods of variability analysis
- Autonomic cardiac control in animal models of cardiovascular diseases II. Variability analysis in transgenic rats with α-tropomyosin mutations Asp175Asn and Glu180Gly
- Fetal ECG extraction during labor using an adaptive maternal beat subtraction technique
- Heart rate variability in the fetus: a comparison of measures
- Estimation of spontaneous baroreflex sensitivity using transfer function analysis: effects of positive pressure ventilation
- Mobile nocturnal long-term monitoring of wheezing and cough
- Vigilance monitoring – review and practical aspects
- Coupled oscillators for modeling and analysis of EEG/MEG oscillations
- Auditory evoked potentials for the assessment of depth of anaesthesia: different configurations of artefact detection algorithms
- NeuMonD: a tool for the development of new indicators of anaesthetic effect
- Recording of focal direct current (DC) changes in the human cerebral cortex using refined non-invasive DC-EEG methodology
- Comparing a template approach and complex bandpass filtering for single-trial analysis of auditory evoked M100
- Wavelet-based analysis of MMN responses in children
- Branched EMG electrodes for stable and selective recording of single motor unit potentials in humans
- EMG analysis of the thenar muscles as a model for EMG-triggered larynx stimulation
- Physiological MR signal variations within the brain at 3 T
- Application of decorrelation-independent component analysis to biomagnetic multi-channel measurements
- A method for locating gradual changes in time series
- The use of digital signal processors (DSPs) in real-time processing of multi-parametric bioelectronic signals
- Steps towards a miniaturized, robust and autonomous measurement device for the long-term monitoring of patient activity: ActiBelt
- Motor timing and more – additional options using advanced registration and evaluation of tapping data
- Cellular signaling: aspects for tumor diagnosis and therapy
- List of reviewers engaged in the Special Issues on Biosignal Processing
- Stellungnahme zu „In vitro Langzeitkultur von humanem Knochen unter physiologischen Lastbedingungen“; Biomed Tech 2004; 49: 364–367
Articles in the same Issue
- Ralph Mueller and Herbert Witte join the Associate Editor team of Biomedizinische Technik/Biomedical Engineering
- Technological innovations in information engineering demand sustained updating and upgrading in biosignal processing applications: a continual renaissance
- Predicting initiation and termination of atrial fibrillation from the ECG
- Predicting the QRS complex and detecting small changes using principal component analysis
- The role of independent component analysis in the signal processing of ECG recordings
- Implantable cardioverter defibrillator algorithms: status review in terms of computational cost
- Assessment of dynamic changes in cerebral autoregulation
- Corrected body surface potential mapping
- Autonomic cardiac control in animal models of cardiovascular diseases. I. Methods of variability analysis
- Autonomic cardiac control in animal models of cardiovascular diseases II. Variability analysis in transgenic rats with α-tropomyosin mutations Asp175Asn and Glu180Gly
- Fetal ECG extraction during labor using an adaptive maternal beat subtraction technique
- Heart rate variability in the fetus: a comparison of measures
- Estimation of spontaneous baroreflex sensitivity using transfer function analysis: effects of positive pressure ventilation
- Mobile nocturnal long-term monitoring of wheezing and cough
- Vigilance monitoring – review and practical aspects
- Coupled oscillators for modeling and analysis of EEG/MEG oscillations
- Auditory evoked potentials for the assessment of depth of anaesthesia: different configurations of artefact detection algorithms
- NeuMonD: a tool for the development of new indicators of anaesthetic effect
- Recording of focal direct current (DC) changes in the human cerebral cortex using refined non-invasive DC-EEG methodology
- Comparing a template approach and complex bandpass filtering for single-trial analysis of auditory evoked M100
- Wavelet-based analysis of MMN responses in children
- Branched EMG electrodes for stable and selective recording of single motor unit potentials in humans
- EMG analysis of the thenar muscles as a model for EMG-triggered larynx stimulation
- Physiological MR signal variations within the brain at 3 T
- Application of decorrelation-independent component analysis to biomagnetic multi-channel measurements
- A method for locating gradual changes in time series
- The use of digital signal processors (DSPs) in real-time processing of multi-parametric bioelectronic signals
- Steps towards a miniaturized, robust and autonomous measurement device for the long-term monitoring of patient activity: ActiBelt
- Motor timing and more – additional options using advanced registration and evaluation of tapping data
- Cellular signaling: aspects for tumor diagnosis and therapy
- List of reviewers engaged in the Special Issues on Biosignal Processing
- Stellungnahme zu „In vitro Langzeitkultur von humanem Knochen unter physiologischen Lastbedingungen“; Biomed Tech 2004; 49: 364–367
