Evaluation of time and frequency domain characteristics of heart rate variability with artificial neural network

  • Authors:

  • Manne Hannula;Esko Alasaarela

  • Affiliations:

  • Medical Engineering R&D Center, Oulu University of Applied Sciences, Finland;Department of Electrical and Information Engineering, University of Oulu, Finland

  • Venue:
  • AIAP'07 Proceedings of the 25th conference on Proceedings of the 25th IASTED International Multi-Conference: artificial intelligence and applications
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

There are several fixed parameters which are in common use in interpretation of heart rate variability (HRV) data. The most common time domain parameters are mean and standard deviation of the R-R intervals, and most common frequency domain parameters are intensities of very low, low and high frequency bands of power spectrum of the R-R intervals. In many studies those parameters are used as input parameters for models which are used for special tasks for example in assessment of autonomic nervous system activity. In this study applicability of artificial neural networks (ANN) for calculation of typical HRV parameters from original R-R interval data was investigated. The analysis was done with help of 24-h ECG recordings of 5 healthy subjects. The results illustrate that by ANN analysis the time domain parameters can be estimated with satisfactory accuracy (R=1.00, R=0.72), and the frequency domain parameters can be estimated with slightly lower accuracy (R=0.51, R=0.49, R=1.00), correspondingly.