Continuous Quantification of Baroreflex and Respiratory Control of Heart Rate by Use of Bivariate Autoregressive Techniques

Background: Normal cardiovascular control involves nonstationary complex interactions between a variety of variables such as heart rate (HR), arterial blood pressure (ABP), and respiratory activity. Methods: To account for both the complexity and transient nature of these phenomena, a closedloop bivariate and time-variant (moving window) model was implemented using autoregressive parametric techniques to identify the typical HR and ABP spectral parameters of low frequency power (LF, 0.03-0.15 Hz), high frequency power (HF, 0.15-0.45 Hz), and their ratio LF/HF. In addition, cross-parameters, such as the gain, phase, and coherent power, between HR, ABP, and changes in instantaneous lung volume (ILV) were computed in both the LF and HF regions. Results: The cross-relations included the HR baroreflex (ABP-HR, alpha), respiratory sinus arrhythmia (ILV-HR), the mechanical influence of respiration (ILV-ABP), and the mechanical feedforward of HR (HR-ABP, beta). The analyses were performed on data from a gradual tilt protocol, which simulates the physiological nonstationarities encountered in daily life. Conclusions: The results were similar to those obtained using a bivariate batch (nonmoving window) Levinson-Wiggins-Robinson algorithm, but the time-variant technique was able to provide nearly continuous parameters, allowing for a real-time continuous monitoring of circulatory control.