Accuracy of some algorithms to determine the oscillometric mean arterial pressure: a theoretical study.

OBJECTIVE: To compare different algorithms to determine the oscillometric mean arterial pressure. METHODS: Using a computer-based model, the accuracy of five algorithms was studied: maximum amplitude algorithm, 33 and 40% formulas calculating the mean arterial pressure from oscillometrically measured systolic and diastolic pressures, and two oscillation shape-based algorithms. We examined to which extent the tested algorithms were influenced by variations in four affective factors: pulse pressure, arterial pressure pulse shape index, and two shape indices (symmetry and steepness) of the artery-cuff pressure/volume relationship. Different ranges of variation of affecting factors were applied. RESULTS: The accuracy of the oscillation shape-based algorithms was found to be higher than the accuracy of the oscillation amplitude-based algorithms. The oscillation mean shape index-based determination had an almost 2-3 times narrower error range compared with the maximum amplitude algorithm. Modeling showed that the mean arterial pressure changes resulting from the varying shape of the arterial pressure waveform cannot be measured using the oscillation amplitude-based algorithms, whereas these changes can be determined effectively using the oscillation shape-based algorithms. CONCLUSION: The maximum amplitude algorithm has a relatively low accuracy for the estimation of the mean arterial pressure. Its error range is even larger than that of estimates calculated by the 33 or 40% formulas from the oscillometrically measured systolic and diastolic blood pressures. A considerably higher accuracy can be achieved by applying the oscillation shape-based algorithms.

Errors of oscillometric blood pressure measurement as predicted by simulation.

OBJECTIVE: To study factors affecting the accuracy of oscillometric measurement. METHODS: By means of a computer-based simulator, a variety of arterial pressure pulses of different shapes and amplitudes were composed as an input signal of the model, whereas cuff volume oscillations were obtained as an output signal. The shape of the artery-cuff pressure/volume relationship was modified. Thereafter, oscillation envelopes were drawn and an estimation of systolic and diastolic pressures was performed by implementing fixed characteristic ratios. The mean arterial pressure was estimated using the maximum oscillation criterion. Altogether, 32 combinations of four affecting factors were studied. RESULTS: For the studied range of affecting factors, the induced errors in systolic pressure, diastolic pressure, and mean arterial pressures were 15, 14, and 27 mmHg, respectively. Systolic readings moved toward underestimation and diastolic readings moved toward overestimation if pulse pressure increased. Arterial stiffening induced systematic overestimation of the systolic pressure and overestimation or underestimation of the diastolic pressure compared with a normal artery, with errors depending on the interaction of the symmetry and steepness indices of the artery-cuff pressure/volume curve. Errors of mean arterial pressure were proportional to pulse pressure, showing overestimation if stiffness increased and/or arterial pressure pulses became steeper. CONCLUSION: Oscillometric readings of systolic and diastolic pressures are strongly influenced by pulse pressure and the shape of the artery-cuff pressure/volume curve, whereas those of the mean arterial pressure are affected by pulse pressure and both the shape of the artery-cuff pressure/volume curve and the shape of the arterial pulse.

A comparison of differential oscillometric device with invasive mean arterial blood pressure monitoring in intensive care patients.

Non-invasive beat-to-beat mean arterial pressure (MAP) in finger arteries recorded by the differential oscillometric device was compared with MAP recorded invasively from A. radialis in 22 patients after cardiac surgery. Based on all 132 paired measurements, the MAP values measured at the radial artery were 2.7 ± 4.9 mmHg higher than those measured on fingers. Among 22 patients there were 8 patients receiving inotropic support, their difference being 2.1 ± 5.6 mmHg. The present study revealed that the mean discrepancy between the invasive radial pressure and finger pressure was small; however, patient data sets showed marked variability in average pressure differences when examined individually.

Influence of pulse pressure variation on the results of local arterial compliance measurement: A computer simulation study.

A computer simulation has been performed to study the influence of the pulse pressure variation on the arterial compliance readings in regard to different calculation techniques and arterial wall elastic properties. We applied the derivative- and amplitude-based (delta-based) calculation techniques to the model of the pressure vs. arterial lumen area relationship of different arteries. The simulated pulse pressure increase resulted in an essential reduction of the delta-based compliance in its near maximum region, and in an increase or no change outside this region. In the case of the relationship of a lower steepness the alterations were smaller.

Theoretical comparison between the model-based and amplitude-based compliance estimation techniques.

A theoretical study has been performed to assess how the pulse pressure variation affects results of the arterial compliance measurement. Two asymmetric arctangent models for the arterial pressure-volume (P-V) relationship have been used by applying model-based and amplitude-based compliance estimation techniques. The results demonstrate that for near-zero transmural pressure the model-based values of the compliance are systematically higher than those estimated by the amplitude-based calculation technique. Increasing of the pulse pressure amplitude causes a considerable reduction in the amplitude-based compliance estimate. For a P-V relationship with a lower steepness, a reduction in the amplitude-based compliance versus transmural pressure is smaller.

Calibrated photoplethysmographic estimation of digital pulse volume and arterial compliance.

Digital pulse volume and arterial compliance have been estimated by using the finger arterial pressure and volume waveforms when performing measurements at rest and during local cold exposure in 17 volunteers. The amplitude-based algorithm determines compliance as a ratio of finger volume and pressure pulses. Cooling was conducted by immersion of the contra-lateral arm in water at 11 degrees C for 2 min. Continuous non-invasive finger blood pressure was measured by the Finapres monitor, while the finger volume pulses were recorded by the UT9201 physiograph, which provides a photoplethysmographic registration of volume changes in a large range of transmural pressures (P(transm)). The photoplethysmographic signal was calibrated by applying a graded modulation to the cuff. The study demonstrated that contra-lateral arm cooling induced a statistically significant (P<0.001) reduction in the digital arterial compliance at P(transm) approximately 0 as well as at P(transm) = 50 mmHg. The median compliance diminished during cooling from 0.67 mm(3 )mmHg(-1) to the level of 55% at P(transm) approximately 0, and from 0.14 mm(3 )mmHg(-1) to the level of 42% at P(transm) = 50 mmHg respectively. The observed changes were mostly due to the reduction in the pulse volume. Digital arterial pressure and volume waveforms recorded by the Finapres monitor combined with a calibrated photoplethysmograph can be used in estimating the absolute values of the dynamic compliance of finger arteries. A quantitative estimation instead of the non-calibrated recording of the digital pulse volume allows shortening the variability and widening the scope of finger arterial compliance measurements.

Asymmetric time-dependent model for the dynamic finger arterial pressure-volume relationship.

The volumetric pulses in the finger PPG signal appear as an information source for several indirect measurement methods. In this study we modelled transformation of pressure pulses into volume pulses in a wide transmural pressure range. The noninvasive finger arterial pressure and photoplethysmographic (PPG) signals were simultaneously registered in 13 healthy subjects while the pressure in the PPG cuff ramped up and down. The nonlinearity of the pressure-volume (P-V) relationship was modelled by an asymmetric function, consisting of two arctangents, each for a different pressure region. The time dependency was described by the first order lag. The disturbing effect of slow creeps in the PPG signal was suppressed by an equal filtering of the measured and model-predicted signal. The differences between the two estimates of the subject's P-V relationship for the increasing and decreasing cuff pressure were small thus showing the repeatability of this method, which can be used for the characterization of individual finger arterial behaviour as well as its changes.

Recording of dynamic arterial compliance changes during hand elevation.

Finger arterial compliance has been studied on the beat-to-beat basis by using the digital arterial pressure and volume waveforms and performing measurements at zero transmural pressure during arm elevation in 11 volunteers. Continuous non-invasive finger blood pressure was measured by applying the Finapres monitor and the finger volume pulses were recorded by the UT9201 physiograph by using the photoplethysmographic principle of registration. Estimation of the beat-to-beat finger arterial compliance is based on (i) the recorded volume and pressure wave amplitudes (Vpulse and Ppulse) and (ii) on the calculation of the slope of the pressure-volume relationship from the first derivatives dV/dt and dP/dt of the recorded volume and pressure pulses near the point of the maximum slope. The results of the study demonstrate that the applied two methods similarly (correlation coefficient r = 0.97) describe the changes of the beat-to-beat compliance during hand elevation test. At the same time the second estimate was 18% higher than the first one (P = 0.003).

Effect of deep breathing test on finger blood pressure.

BACKGROUND: The deep breathing test (DB) is a simple method to measure heart rate variability. However, in most studies the amplitude of blood pressure oscillations has not been considered and little is known about the influence of the deep breathing test on the blood pressure level. DESIGN: The aim of this study is to investigate the effect of DB on finger blood pressure (FBP). METHODS: Continuous beat-to-beat FBP was recorded by the volume clamp method (Portapres model 2 monitor). RESULTS: Thirteen volunteers were studied before, during and after DB at a fixed rate of six breaths/min. After DB, systolic blood pressure (SBP), diastolic blood pressure (DBP) and mean blood pressure (MBP) were lower than before DB by 9.1, 4.3 and 5.9 mmHg, respectively. There was no significant correlation between the reduction in the BP level and the amplitudes of induced oscillations in blood pressure and heart rate during the deep breathing test. CONCLUSIONS: Deep breathing might be used to reduce blood pressure.

Beat-to-beat measurement of the finger arterial pressure pulse shape index at rest and during exercise.

Arterial pressure waveform can be characterized by the pulse shape index kpulse determined as kpulse = (Pmean - Pdiast) / (Psyst - Pdiast). For brachial artery, the shape index value of 0.33 is usually applied to approximate Pmean from the measured Psyst and Pdiast. Our purpose was to test whether this value can validly be applied to finger vascular beds under different experimental conditions. By using Finapres, we non-invasively estimated the beat-to-beat values of kpulse in the fingers of young healthy persons in supine position at rest and during a 4-min moderate exercise (rhythmical exercise with the quadricep muscles in combination with handgrip compression). To detect intensive peripheral vasoconstrictions, a laser-Doppler probe was attached to the thumb pulp of the same hand. Periods of 30 s without intensive vasoconstriction for rest, different stages of exercise and recovery were involved in the analysis in every subject. The results demonstrated that the group-averaged value of kpulse (median with a 95% confidence interval) in the fingers of 11 healthy volunteers aged from 20 to 24, equalled 0.33 (0.31- 0.34), 0.31 (0.28-0.34), 0.35 (0.33-0.39) and 0.38 (0.34-0.43) for rest, first and second stages of exercise and recovery, respectively. We conclude that in the fingers of young healthy persons in supine position formula Pmean = Pdiast + 1/3 (Psyst - Pdiast) gives an adequate approximation for rest and low intensity exercise (first stage), and slightly underestimates the actual finger mean blood pressure during moderate exercise (second stage) and recovery.

Portapres and differential oscillometric finger blood pressure changes during deep breathing test in the assessment of BRS index.

The aim of this study was to investigate changes in the finger blood pressure during a deep breathing test (DB) and to find out whether the mean blood pressure might be used as a substitute for the systolic pressure in calculations of baroreflex sensitivity from data derived from the DB test. Continuous beat-to-beat finger pressure was recorded by the volume clamp method (Portapres model 2 monitor). In addition, the mean arterial pressure was recorded by the modified oscillometric method (UT9201 beat-to-beat monitor, University of Tartu, Estonia). Fifteen healthy volunteers, aged 25-56 years, were studied. The amplitude of respiration-linked oscillations in the Portapres systolic (Psyst) and mean blood pressure (Pmean) was 22.2 +/- 8.8 and 16.6 +/- 5.8 mmHg, respectively. There was no significant difference between the amplitudes of induced changes in Pmean recorded by the two devices: the amplitude of oscillations in the mean blood pressure recorded by the differential oscillometric monitor was 16.0 +/- 5.9 mmHg. The amplitude of oscillations in Psyst correlated significantly with the amplitude of oscillations in Pmean recorded either by Portapres or by UT9201 (r=0.95 and 0.98, respectively). The high correlation between the amplitudes of oscillations in mean and systolic blood pressure allows to conclude that mean arterial pressure changes during a DB test might be used instead of systolic pressure changes in calculation of the ratio of changes in pulse interval to changes in blood pressure, which is considered to reflect baroreflex sensitivity.

Simultaneous recording of fingertip skin blood flow changes by multiprobe laser Doppler flowmetry and frequency-corrected thermal clearance.

We compared the results of skin blood flow (SBF) measurements, obtained simultaneously in adjacent fingertips by laser Doppler flowmetry (LDF) and thermal clearance (TC) probes, having approximately the same spatial and temporal characteristics. Experiments were performed in nine healthy volunteers during rest at room temperature (26-28 degrees C). A time resolution equal to about a second was achieved by speeding up the response of the thermal measurement circuit by applying the computer simulation software of the MATLAB package. The comparison revealed that the frequency-corrected TC signal correlated well with the multiprobe LDF signal (median correlation coefficient = 0.90, range = 0.84 to 0.96). At the same time the individual slope values of the regression equation ranged from 0.58 to 1.61, revealing the difficulties encountered in obtaining the invariant scaling factor between the TC and LDF measurements. The relationship between the frequency-corrected TC signal and the multiprobe LDF signal was found to be linear in the range of SBF changes of about three- to fourfold. In the case of larger fluctuations in SBF, excessive acceleration at high SBF rates was noted.