Time Series Calculation of Heart Rate Using Multi Rate FIR Filters

RISK, MARCELO; FERNÁNDEZ SLEZAK, DIEGO; TURJANSKI, PABLO; PANELLI, ALEJANDRO; TABORDA, RAM; MARSHALL, GUILLERMO

EEUU

Congreso; Computers in Cardiology 2007; 2007

Computers in Cardiology

The spectral analysis of heart rate variability, based<br />on the Fourier transform, needs even sampled data. The<br />objectives of this study were to develop an interpolation<br />method based on multi rate FIR filters, and then to imple-<br />ment this method for parallel processing machines. A total<br />of three data sets were used: a) simulated heart rate with<br />an IPFM model, b) autonomic blockage database (both<br />pharmacological and postural), and c) long term Holter<br />studies (recordings of 24 hours). Spectral analysis, for the<br />three data sets, was processed for both interpolation using<br />FIR filters and cubic splines, the results for Bland and Alt-<br />man analysis for low frequency band, showed a difference<br />of -47±131 ms2 ; then for the high frequency band, the<br />difference was 3±48 ms2 . The presented method of time<br />series calculation, using FIR filters, probed to be equiva-<br />lent for both simulated and real data, and is suitable for<br />parallel programming implementation.<br /><br />