Pre-Processing of Anomalies in Telemetry Data for Heart Rate Variability Analysis

Kline JL, Dabisch PA.

Presenter of poster topic presented as part of the National Cancer Institute’s Spring Research Festival, Frederick, MD

Abstract:
While electrocardiogram measurements of animals is nothing new, only recently has implantable telemetry devices made long term observation practical. Thirty days of periodic data collection is not unheard of, and this expanded dataset gives rise to the potential for more meaningful post-challenge feedback. 

The implants used in this study utilize a two-lead  type interface. The largest signal amplitude obtained, termed the R wave, is used to calculate heart rate. An R-wave corresponds to the cardiac cell depolarization that triggers contraction of the left and right ventricles of the heart. The interval time between this peak and the next successive  beat is the inverse of the heart rate. The variability in the R-R interval is referred to as heart rate variability (HRV) and arises due to fluctuations in the activity of the nervous systems responsible for the control of the heart rate. 

Noise is always an issue with physiological measurements, and this case was no exception. Any number of factors can influence the amount of noise present in an EKG signal, and the level of noise is not always consistent (due to changes in animal posture, activity level, etc). The acquisition software was not always able to reliably detect all QRS complexes within the EKG signal. When this happens, a either an abnormally long R-R interval or a gap is left in the reported data. The results of HRV analysis, especially frequency domain or spectral analysis, can be significantly influenced by the presence of these longer R-R interval and gaps.

The goal of this study was to compensate for these long R-R and missed beat calculations in the acquisition software. To achieve this a method was introduced using LabVIEW to detect abnormal R-R segments in the EKG and fill in for one or two beats. The data use to test this process, was provided from two African Green monkeys.