Brno University of Technology ECG Signal Database with Annotations of P Wave (BUT PDB)

Abstract

Brno University of Technology ECG Signal Database with Annotations of P Wave (BUT PDB) is an ECG signal database with marked peaks of P waves created by the cardiology team at the Department of Biomedical Engineering, Brno University of Technology. The database consists of 50 2-minute 2-lead ECG signal records with various types of pathology. The ECGs were selected from three existing databases of ECG signal - the MIT-BIH Arrhythmia Database, the MIT-BIH Supraventricular Arrhythmia Database, and the Long Term AF Database. The P waves positions were manually annotated by two ECG experts in all 50 records. Each record contains also annotation of positions and types of QRS complexes (from original database) and dominant diagnosis (pathology) present in record. This database is created for the development, evaluation and objective comparison of P wave detection algorithms.

Background

Accurate detection of P waves and subsequent cardiac pathological events is an important part of electrocardiogram (ECG) evaluation. Currently, methods for P wave detection in physiological conditions are well-described and efficient. However, results of testing on pathological signals are insufficient. A key reason for this is the limited availability of public ECG databases with annotated P waves and pathologies. There are 3 publicly available databases with manually annotated P waves - the QT database [1-3] (contains only physiological signals), MIT-BIH Arrhythmia Database P-Wave Annotations [4-6] (contains only few types of pathologies) and the Lobachevsky University Electrocardiography Database [7]. Here we introduce a new database with annotated P waves in signals with 23 different types of pathology.

Methods

The ECGs were selected from three existing databases of ECG signal - the MIT-BIH Arrhythmia Database [8], the MIT-BIH Supraventricular Arrhythmia Database [9,10], and the Long Term AF Database [11, 12]. Two-minute sections of these records were selected by ECG experts, who sought to identify "interesting" pathological signals. The database therefore has a higher incidence of pathologies than typically expected.

The P waves positions were manually annotated by two ECG experts for each beat in all 50 records. The first expert provided manual annotations, and the second checked them. Unclear parts of records were checked by both experts until a consensus was reached. These tasks were conducted manually, without the use of automated software. To facilitate the work of the ECG experts, a free software tool, SignalPlant [13], was used for manual annotation of P waves.

Data Description

The database consists of 50 2-minute 2-lead ECG signals with various types of pathology with annotated P waves, selected from 3 existing databases of ECG signals [8, 9, 11]. The P waves were manually annotated by two ECG experts for each beat in all 50 records. Each record also contains annotation of positions and types of QRS complexes (from original databases) and dominant diagnosis (pathology).

In BUT PDB, there are 5437 P waves, 7638 QRS complexes of which 2201 are without P wave. In this database, 23 different types of pathology are present. Types of pathologies with their abbreviations used in database are listed in Table below. The exact types of pathologies in each signal are described in the  "README".

All data are provided in the WaveForm Database (WFDB) format. The names (IDs) of the recordings are numbers from 01 to 50. The ECG signals are in files: *.dat, *.hea, and the annotations of P waves are in files with names *.pwave, the positions of QRS complexes and their types and sampling frequency of signal are in files with name *.qrs

Usage Notes

There are few publicly available databases of ECG signals that are accompanied by well-curated annotations of P waves. Our goal in creating the dataset was to help address this issue. The database may be be valuable for the development, evaluation and objective comparison of P wave detection algorithms.

Acknowledgements

This work was funded by the United States Office of Naval Research (ONR) Global, award number N62909-19-1-2006.

Conflicts of Interest

The authors declare that there are no known conflicts of interest.

References

1. Goldberger, A., Amaral, L., Glass, L., Hausdorff, J., Ivanov, P. C., Mark, R., ... & Stanley, H. E. (2000). PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals. Circulation. 101(23). http://doi.org/10.1161/01.CIR.101.23.e215
2. Laguna, P., Mark, R. G., Goldberg, A., & Moody, G. B. (1997). A database for evaluation of algorithms for measurement of QT and other waveform intervals in the ECG. Computers in cardiology 1997 (pp. 673-676)
3. Laguna P, Mark RG, Goldberger AL, Moody GB. (1999). QT Database (version 1.0.0). PhysioNet. https://doi.org/10.13026/C24K53
4. Maršánová L, Němcová A, Smíšek R, Goldmann T, Vítek M, Smital L. (2018) MIT-BIH Arrhythmia Database P-Wave Annotations (version 1.0.0). PhysioNet. https://doi.org/10.13026/C2108F
5. Moody, G. B., & Mark, R. G. (2001). The impact of the MIT-BIH arrhythmia database. IEEE Engineering in Medicine and Biology Magazine, 20(3), 45-50.
6. Maršánová, L., Němcová, A., Smíšek, R., Goldmann, T., Vítek, M., Smital, L. Automatic Detection of P Wave in ECG During Ventricular Extrasystoles. World Congress on Medical Physics and Biomedical Engineering 2018, Singapore, 2018. (pp. 381-385)
7. Kalyakulina, A., Yusipov, I., Moskalenko, V., Nikolskiy, A., Kozlov, A., Kosonogov, K., Zolotykh, N., & Ivanchenko, M. (2020). Lobachevsky University Electrocardiography Database (version 1.0.0). PhysioNet. https://doi.org/10.13026/qweb-sr17
8. Moody GB, Mark RG. (2005) MIT-BIH Arrhythmia Database (version 1.0.0). PhysioNet. https://doi.org/10.13026/C2F305
9. Greenwald SD. (1999) MIT-BIH Supraventricular Arrhythmia Database (version 1.0.0). PhysioNet. https://doi.org/10.13026/C2V30W
10. Greenwald SD. Improved detection and classification of arrhythmias in noise-corrupted electrocardiograms using contextual information (1990). PhD thesis, Harvard-MIT Division of Health Sciences and Technology.
11. Petrutiu S, Sahakian AV, Swiryn S. (2008) Long Term AF Database (version 1.0.0). PhysioNet. https://doi.org/10.13026/C2QG6Q
12. Petrutiu, S., Sahakian, AV., Swiryn, S. Abrupt changes in fibrillatory wave characteristics at the termination of paroxysmal atrial fibrillation in humans (2007). Europace 9:466-470. http://doi.org/10.1093/europace/eum096
13. Plesinger, F., Jurco, J., Halamek, J., & Jurak, P. (2016). SignalPlant: An open signal processing software platform. Physiological Measurement, 37(7). http://doi.org/10.1088/0967-3334/37/7/n38