Reliability of transfer function estimates in cardiovascular variability analysis

Abstract

Transfer function (TF) analysis is a widely diffused technique in the assessment of the relationship between short-term cardiovascular variability signals, particularly blood pressure, heart rate and respiration. To guarantee the reliability of the estimates, a conventional threshold of 0.5 on the magnitude squared coherence (MSC) is commonly used, although (i) other analysis parameters play a role and (ii) lower values of MSC are frequently unavoidable in physiological systems. In this study, computer simulations are performed to assess the dependency of the bias and standard deviation (SD) of TF estimates on record length (RL), spectral window bandwidth (Bw) and MSC; to evaluate the accuracy of theoretical expressions for the computation of the confidence interval (CI) of the estimates; and to assess, in some representative situations, how faithfully observed TF shapes reproduce the underlying true functions in conditions of very low MSC. The accuracy of TF estimates increases non-linearly with increasing RL, and the benefit over 7 min is small. Using this RL, the relative bias for the TF modulus is < 10% for MSC > 0.2. Estimates of TF phase are unbiased. The SD of both the modulus and phase increases linearly as the MSC decrease to 0.4 and then, for lower MSC, increases markedly with nonlinear behaviour. Bw= 0.03Hz appears to be most suitable to reduce the error, preserving spectral resolution. CIs for the TF phase are highly reliable, whereas those for the modulus tend to be slightly narrower than the nominal value at high coherence values. Major features of the TF shape appear to be preserved in simulations with very low MSC. The major problem in TF estimation is the sharp increase in the variability of the measurements as the coherence decreases towards the lowest values. The combination of RL > or = 420s and Bw= 0.03Hz should be suggested in short-term cardiovascular variability studies. Although basic features of the true TF can be recovered even when the MSC is < 0.5, much greater values can be necessary when accurate point estimates are needed. Theoretical expressions for the computation of confidence intervals of the TF are adequate for practical purposes.

Assessment of cerebral autoregulation: the quandary of quantification

“…This approach is equivalent to setting the threshold at zero. This approach produces phase and gain estimates that are comparable with those obtained under conditions of high coherence (e.g., during cyclic squat-to-stand exercise) (8,53), presumably because averaging filters out the random variability of the function around its true value (38).…”

Different Predictive Values of Electrophysiological Testing and Autonomic Assessment in Patients Surviving a Sustained Arrhythmic Episode

“…Recently, however, this approach has been criticized on the grounds that (i) the reliability of the estimates depends on other parameters besides the coherence and (ii) low values of coherence are unavoidable in physiological systems owing to an impaired baroreflex brought about by a pathological condition and/or a low signal-to-noise ratio in the signals. 23 Indeed, it has been shown that the measurability of BRS with the Robbe method in post-MI, chronic heart failure and geriatric patients the measurability is as low as <50%. 15,24 In the present study this index could be measured in 65 of 112 patients (58%).…”

Early autonomic dysfunction in type 1 diabetes: a reversible disorder?

“…BRS was determined from spontaneous fluctuations in the RR interval and SBP during the 15 and 6 breaths per minute recordings using the sequence method (BRS+/+, BRS−/−) [34], the alpha coefficient (Alpha low-frequency BRS) [35] and the transfer function technique (TF-BRS) [36]. In the sequence method, BRS was estimated by identifying spontaneously occurring sequences of three or more consecutive heartbeats in which both SBP and the subsequent RR intervals changed in the same direction.…”