In this paper, we developed and evaluated a robust single-lead electrocardiogram (ECG) delineation system based on the wavelet transform (WT). In a first step, QRS complexes are detected. Then, each QRS is delineated by detecting and identifying the peaks of the individual waves, as well as the complex onset and end. Finally, the determination of P and T wave peaks, onsets and ends is performed. We evaluated the algorithm on several manually annotated databases, such as MIT-BIH Arrhythmia, QT, European ST-T and CSE databases, developed for validation purposes. The QRS detector obtained a sensitivity of Se=99.66% and a positive predictivity of P+=99.56% over the first lead of the validation databases (more than 980,000 beats), while for the well-known MIT-BIH Arrhythmia Database, Se and P+ over 99.8% were attained. As for the delineation of the ECG waves, the mean and standard deviation of the differences between the automatic and manual annotations were computed. The mean error obtained with the WT approach was found not to exceed one sampling interval, while the standard deviations were around the accepted tolerances between expert physicians, outperforming the results of other well known algorithms, especially in determining the end of T wave.

In this paper we assessed the possibility of using the pulse rate variability (PRV) extracted from the photoplethysmography signal as an alternative measurement of the HRV signal in non-stationary conditions. The study is based on analysis of the changes observed during a tilt table test in the heart rate modulation of 17 young subjects. First, the classical indices of HRV analysis were compared to the indices from PRV in intervals where stationarity was assumed. Second, the time-varying spectral properties of both signals were compared by time-frequency (TF) and TF coherence analysis. Third, the effect of replacing PRV with HRV in the assessment of the changes of the autonomic modulation of the heart rate was considered. Time-invariant HRV and PRV indices showed no statistically significant differences (p > 0.05) and high correlation (>0.97). Time-frequency analysis revealed that the TF spectra of both signals were highly correlated (0.99 +/- 0.01); the difference between the instantaneous power, in the LF and HF bands, obtained from HRV and PRV was small (<10(-3) s(-2)) and their temporal patterns were highly correlated (0.98 +/- 0.04 and 0.95 +/- 0.06 in the LF and HF bands, respectively) and TF coherence in the LF and HF bands was high (0.97 +/- 0.04 and 0.89 +/- 0.08, respectively). Finally, the instantaneous power in the LF band was observed to significantly increase during head-up tilt by both HRV and PRV analysis. These results suggest that although some differences in the time-varying spectral indices extracted from HRV and PRV exist, mainly in the HF band associated with respiration, PRV could be used as a surrogate of HRV during non-stationary conditions, at least during the tilt table test.

We present here a QT database designed jbr evaluation of algorithms that detect waveform boundaries in the EGG. The dataabase consists of 105 fifteen-minute excerpts of two- channel ECG Holter recordings, chosen to include a broad va- riety of QRS and ST- T morphologies. Waveform bounda,ries for a subset of beats in, these recordings have been manually determined by expert annotators using an interactive graphic disp1a.y to view both signals simultaneously and to insert the annotations. Examples of each m,orvhologg were inchded in this subset of uniaotated beats; at least 30 beats in each record, 3622 beats in all, were manually a:anotated in Ihe database. In 11 records, two indepen,dent sets of ennotations have been inchded, to a.llow inter-observer variability slwdies. The Q T Databnse is available on a CD-ROM in the format previously used for the MIT-BJH Arrhythmia Database ayad the Euro- pean ST-T Database, from which some of the recordings in the &T Database have been obtained.

This paper reviews the current status of principal component analysis in the area of ECG signal processing. The fundamentals of PCA are briefly described and the relationship between PCA and Karhunen-Loève transform is explained. Aspects on PCA related to data with temporal and spatial correlations are considered as adaptive estimation of principal components is. Several ECG applications are reviewed where PCA techniques have been successfully employed, including data compression, ST-T segment analysis for the detection of myocardial ischemia and abnormalities in ventricular repolarization, extraction of atrial fibrillatory waves for detailed characterization of atrial fibrillation, and analysis of body surface potential maps.