System for automatic heart rate calculation in epileptic seizures

Figures

• 

  Fig. 1. A typical ECG waveform with P, Q, R, S and T waves

• 

  Fig. 2. A typical ECG and EEG signals immediately prior to and during epileptic seizure—from top: EEG (FP1, F3, P3) and ECG

• 

  Fig. 3. Diagram of the system for determining the RR intervals

• 

  Fig. 4. Raw ECG signal

• 

  Fig. 5. A registered ECG signal after band-pass (4–80 Hz) and band-stop (48–52 Hz) filtering and normalization

• 

  Fig. 6. A sample reference QRS complex (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{\Psi }}_{QRS}$$\end{document}ΨQRS)

• 

  Fig. 7. A typical changes of HR value before and during epileptic seizure (the vertical blue line shows the beginning of an epileptic seizure)

• 

  Fig. 8. Detection of HR (location of R-waves)

• 

  Fig. 9. ECG waveform, a noise signal of strength k and the ECG signal with added noise

• 

  Fig. 10. R-wave detection accuracy, blue color—the direct method, red color—the correlation method

• 

  Fig. 11. The results of the R-wave detection with the use of Pan & Tompkins (asterisk) and our (open circle) algorithms for the real ECG signal disturbed by natural muscle artifacts

• 

  Fig. 12. The results of the R-wave detection with the use of Pan & Tompkins (asterisk) and our (open circle) algorithms