G. Baselli

Cardiovascular variability signals: towards the identification of a closed-loop model of the neural control mechanisms

The authors consider parametric methods for processing cardiovascular signals and try to provide a global, although indirect evaluation of some neural regulatory activities. In particular, the variability signals of the heart rate (under the form of interval tachogram) and arterial blood pressure (systogram) together with respiratory movement signal (respirogram) are considered as inputs to a closed-loop model which describes a few aspects of the physiological interactions among the signals themselves. The identifiability of the transfer function of the model is demonstrated from the joint process black-box description of the signals. A direct identification procedure is proposed dividing the system into two dynamic adjustment models. A few suggestions are deduced on how and where the respirogram enters the model and on the genesis of the 10-s rhythm, parameters relevant to the Starling effect, Windkessel model, and the gain of baroreceptor mechanisms. The approach presented is intended also to provide a general frame for closed-loop identification in different pathophysiological conditions.<<ETX>>

Spectral decomposition in multichannel recordings based on multivariate parametric identification

A method of spectral decomposition in multichannel recordings is proposed, which represents the results of multivariate (MV) parametric identification in terms of classification and quantification of different oscillating mechanisms. For this purpose, a class of MV dynamic adjustment (MDA) models in which a MV autoregressive (MAR) network of causal interactions is fed by uncorrelated autoregressive (AR) processes is defined. Poles relevant to the MAR network closed-loop interactions (cl-poles) and poles relevant to each AR input are disentangled and accordingly classified. The autospectrum of each channel can be divided into partial spectra each relevant to an input. Each partial spectrum is affected by the cl-poles and by the poles of the corresponding input; consequently, it is decomposed into the relevant components by means of the residual method. Therefore, different oscillating mechanisms, even at similar frequencies, are classified by different poles and quantified by the corresponding components. The structure of MDA models is quite flexible and can be adapted to various sets of available signals and a priori hypotheses about the existing interactions; a graphical layout is proposed that emphasizes the oscillation sources and the corresponding closed-loop interactions. Application examples relevant to cardiovascular variability are briefly illustrated.

Reliability of the measurement of RT variability

The variability of the interval between the R peak and the apex of T wave has a power 2-3 orders of magnitude smaller than that of RR interval variability therefore, adequate recording, pre-processing and detection procedures are required to avoid artefacts. Emphasis is posed on the dramatic differences occurring between different leads, which are explained as artefacts induced by changes in the cardiac electrical axis. A satisfactory compensation appears to be provided by the modulus signal from three orthogonal leads. So, the low frequency (LF) and high frequency (HF) components were clearly recognised in 14 normal (N) subjects and in 12 patients 2 weeks after myocardial infarction (MI). This was confirmed by the cross-spectrum between the RT and the RR series. In N, HF (36.7/spl plusmn/23.1 nu.) was predominant over LF (25.4/spl plusmn/14.5), while in MI the ratio was inverted (LF=37.0/spl plusmn/23.3, HF=25.6/spl plusmn/13.9).<<ETX>>

Beat-to-beat variability of microvascular peripheral resistances assessed with a non-invasive approach

The pressure-flow relationship at the peripheral level is non-invasively studied in human subjects: the impedance function and the beat-to-beat variability series of microvascular peripheral resistance are estimated. The frequency content of this variability signal is compared to those of more classical variability series at rest and during mild supine physical exercise.

A comparison between R-R and R-T variability signals in normal and pathological subjects

Power spectrum analysis of the RR variability (RRV) signal is compared with the spectrum obtained from the RT variability (RTV) signal. The analysis is carried out on a population of 16 normal subjects and 24 patients, 15 days after an acute episode of myocardial infarction (MI). While in normal subjects the RTV signal is satisfactorily correlated with RRV, presenting the same trend in a rest/tilt maneuver, in MI patients such a correlation strongly decreases. This finding demonstrates that the pathophysiological link between the two variability signals is complex, and local properties of the ventricular cells may be accounted for by the noticed difference in their behavior. RRV and RTV signal analysis can be used to investigate the relations between the beat-to-beat variability of ventricular depolarization and repolarization durations.<<ETX>>

Synchronisation analysis of heart period variability signal based on corrected conditional entropy

Regularity and synchronisation between different rhythms mainly at low (LF) and high frequency (HF) is evaluated in heart rate variability through conditional entropy (CE). CE quantifies the degree of information left after regular patterns have been removed. In short (about 300 points) data sequences CE is evaluated with a correction which considers as random the patterns that appear only one time and would otherwise be considered as regular. The minimum of the corrected CE (CCE) with increasing pattern length is chosen as the most reliable CE estimate. Beat-to-beat heart period series were analysed in 8 normal subjects at rest (R), head up tilt (T) and controlled respiration (CR). A significant decrease in CCE was found in T (0.9/spl plusmn/0.052) compared to R (1.01/spl plusmn/0.037); during CR, CCE displayed a slight and not significant increase (1.07/spl plusmn/0.055). The enhanced LF activity present in T seems to have a greater regularising power than the periodical input provided by CR at HF.

The synchrony between baroreflex sequences and cardio-respiratory activity

The bias observed between methods developed to assess the baroreflex control of Heart Rate (HR) lead to gain further insight into those methods. In two groups of subjects were studied: a group of young adults, and a group of middle-aged adults, cardiovascular (CV) and respiratory parameters were continuously recorded during a graded bicycle exercise. This study focused on the sequence method, and its particular time sequence pattern (up and down sequences). Features of respectively the up and down sequences were analysed through comparative analysis and sequence pattern analysis. No significant changes were noticed between the up and down sequences for both the number of sequences and the baroreflex gain. A synchrony was present according to the phase of respiration and also with the diastolic arterial pressure (dap). It may suggest a relation between the baroreflex function and the Bainbridge reflex in governing CV oscillations.

Different interference patterns among spontaneous low and high frequency oscillations and forced ventilation in sympathetic outflow

In the sympathetic outflow directed to the heart of decerebrate artificially ventilated cats, nonlinear interferences among spontaneous low frequency waves, (LF, /spl sim/0.1 Hz) related to vasomotor control, an high frequency (HF, /spl sim/0.3 Hz) from respiratory centers, and forced ventilation (at 0.3 Hz) were investigated. Various methods were tested for the simultaneous processing of sympathetic activity (Sy) and ventilation (V). Time-space separation plot (TSSP) and frequency tracking locus (FTL) were used to detect phase locking (PL) and quasiperiodicity (QP). Deterministic patterns hidden in aperiodic data were also searched for. In comparison to control (n=17), sympathetic activation (inferior vena cava occlusion, n=17) increased PL of LF to V, while the reverse was observed during sympathetic inhibition (aortic constriction, n=13). After spinal section, (n=26) aperiodicity with intermittency was found.

Low and high frequency non-linear interactions in the sympathetic discharge: contribution of spinal centers

When biological oscillators interact as a result of nonlinear mechanisms several phenomena such as period doubling, phase locking, complex and irregular behaviors may occur. These behaviors depend on the degree of the coupling among the rhythmic generators and on the amplitude and frequency of the relevant oscillations. In decerebrate artificially ventilated cats low (around 0.1 Hz) and high (around 0.3 Hz) frequency rhythms present in the preganglionic efferent sympathetic activity directed to the heart are affected by forced periodical ventilation. These interactions result in various ratios between sympathetic discharge and ventilatory cycles (1:1, 1:2, 1:3, 1:4 and lower). After spinal section, as a consequence of the reduction of the coupling between central oscillators, external ventilation and sympathetic activity, irregular behaviors including broad band patterns, weak low frequency rhythms and period-three periodicities are classified.<<ETX>>