外文文献翻译_压电传感器测定动脉脉搏波速度

Piezoelectric sensor determination of arterial pulse wave velocity

Arterial pulse wave velocity (APWV) is a measure of the elasticity (or stiffness) of peripheral arterial blood vessels. The pulse referred to here will be the pressure pulse as opposed to the flow pulse measured by ultrasound Doppler.

The pressure pulse velocity varies over the range from about 12 m s?1 to

15 m s?1 in stiff peripheral arteries, whereas in normal arteries it has a velocity in the range of 7 to 9 m s?1.

The aim of this project was the development of a fast and easy to use system for the determination of peripheral arterial pulse wave velocity. The principle of the PWV measurement is based on simultaneous measurement of two pulse waves at two different positions, such as the radial artery at the wrist and the brachial artery just above the elbow. By determining the pulse transit time between these points and the distance measured between the two locations, pulse wave velocity may then be calculated. The pressure pulse detection is done by using two piezoelectric sensors which generate a measurable voltage at the output contacts if they are mechanically deformed. The deformation produced voltage is first amplified and filtered and then digitalized with a data acquisition card. The analysis of the data obtained from the sensors includes a filtering process, the calculation of the PWV with three different methods— foot-to-foot, cross-correlation and peak-to-peak—and the determination of the arterial pulse rate.

Extensive measurements with human test subjects were carried out to optimize the techniques of data acquisition and analysis. For example, it was found that the best procedure was to hold the sensors in place using elastic straps alone. The data analysis was upgraded with an additional software module, which deletes, in effect, outriders or invalid measurements. With the optimized system, a series involving eight test subjects ranging in age from 22 to 32 years was completed (all normotensive). The arterial pulse wave velocities determined covered a range from 6 m s?1to 12 m s?1, with an average standard deviation of less than 2.5 m s?1for individual results. These are slightly higher,but close to published APWV data. The results showed that reproducible results can be obtained with the existing PWV acquirement and analysis system.

The measurement of arterial pulse wave velocity (APWV) is one of the methods used to measure physiological changes in peripheral vascular disease. Others include pressure pulse contour, arterial elasticity, pulsatile flow, complex vascular impedance and cardiac work.There have been many investigations over the past 30 years or so to relate changes in age and progress of arterial disease, to vessel pathology and distensibility based on the propagation characteristics of the arterial pulse (Malindzak and Meredith 1970, McCormack 1981, Persson et al 2001, Ramsey 1995, Wilkinson 1998). They have only been partially successful from a clinical point of view, mainly due to the difficulty in controlling the effects of patient parameters

(autonomic system, patient movement, etc).Early on, the methods of determining APWV involved direct invasive measurement of the pulse pressure at two points along an arterial segment and measuring the time taken for the pulse to travel the length of the segment. Such investigations had to be limited to animals usually dogs.

The sensor technique used in this work involves the piezoelectric effect in polyvinyllidene fluoride (PVDF), which produces an output voltage in response to mechanical pressure on the material.

Three methods of APWV determination are used: foot-to-foot APWV; peak-to-peak APWV and cross-correlation APWV. The FFAPWV and CCAPWV methods are less sensitive to pressure wave reflections at bifurcations, etc in the arterial tree, than the PPAPWV method. Mean values and standard deviations were computed for all three methods and compared.

‘Foot-to-foot’ APWV (FFAPWV). This is based on the velocity of the ‘foot’, or leading edge, of the pressure pulse wave. The arrival times of the foot of the pulse wave at two positions along the artery are recorded. If Δt is the difference in arrival times and Δs the distance between the two recording positions (proximal and distal), the FFAPWV is simply

FFAPWV = Δs/Δt.

‘Peak-to-peak’ APWV (PPAPWV). This is completely analogous to the

FFAPWV except that the points of observation are the two (proximal and distal) peaks of the pulse wave and

PPAPWV = Δs/Δt.

Apparent pulse wave velocity (AAPWV). The pressure wave may be represented as a Fourier series,

P(t) = P0 +?Pn sin(nωt + θn)

n?1kWhere P0 is the mean fluid pressure, n is the harmonic number, Pn is the amplitude of the nth harmonic and θn is the phase angle of the nth harmonic.

The spatial rate of change of the phase for one harmonic based on two

simultaneous pressure measurements separated by a distance _s along an artery, is related to the apparent arterial pulse wave velocity (AAPWV) by the following equation,

AAPWVn = (Δs)n(f )(360o)/(θx1? θx2)

Where AAPWV n is the apparent pulse wave velocity for the nth harmonic, f is the heart rate,θx1 is the phase angle for the proximal harmonic n and θx2 is the

phase angle for the distal harmonic n.

Cross-correlation PWV. If the arterial pulse at the proximal measurement position is represented by the pressure time series P(x1, t) and that at the distal position by P(x2, t) and

the cross-correlation coefficient is Φx1,x2(τ ), then Φwill have a maximum value at some time lag.

The correlation function can be expressed as

Φx1,x2(τ ) = (1/T )?ε/2?ε/2P(x1, t)P(x2, t) dt.

The value of τ at which maximum correlation occurs represents the transit time (_t) of the pressure wave from position x1 to position x2 along the arterial segment. From the separation distance and transit time data the correlation arterial pulse wave velocity is

CCAPWV = (x2 ? x1)/ Δt.

In this work normal, young test subjects were used, and it has the primary objectives of optimizing the measurement procedures and establishing the statistical spread and mean values of the observed PWVs for a specific peripheral arterial segment. Based on this, it is planned to use the system in clinical trials involving patients with peripheral arterial disease (due to diabetes, hypertension, etc), pre-, during and post treatment (pharmaceutical or surgical). Analogue and digital circuitry

Analogue charge amplifier. Piezoelectric materials convert mechanical stress or strain into proportionate electrical energy, by producing a charge when subjected to mechanical stress. The charge is converted to a voltage by an operational amplifier connected as a current integrator, called a charge amplifier. The signal output of the amplifier is approximately ?30 mV. It is augmented by signal amplification.

Analogue signal amplification . This is done by use of an inverting amplifier. Because a dc signal appears at the output of the charge amplifier, dc offset removal is essential and is implemented in the inverting summing

The next phase of the analogue circuitry is a low pass filter to remove the 50 Hz noise interference.

Digital controlled data acquisition and analysis. A data acquisition board (DAQ) is required when the transducer signals need interfacing with a PC. The board contains 12 bit plus sign and a successive approximation and self-calibrating

analogue-to-digital (ADC) converter. The ADC incurs a systematic error known as the quantization error. It is due to limited resolution and with the analogue input limit set at ?5 to +5 V, the quantization error of the A/D converter used here was calculated to be 0.122 mV.

The data acquisition and analysis was done using Lab View—a powerful instrumentation and analysis programming language for PCs.

Digital data acquisition program. The data acquisition circuit performs all the necessary operations for the data acquisition with Lab View. The functions of the

circuit initialize the data acquisition and read the data from the acquisition card. These data are stored for later use in the data analysis part of the program.

Lab View programs are called virtual instruments (VIs) because of their

appearance and operations are analogous to measuring instruments. A VI that is called within another VI is called a sub-VI and is analogous to a subroutine in text-based languages.

The data acquisition, analysis and presentation are comprised of three main procedures:

(1) Data acquisition card to interface the hardware to the PC.

(2) Data acquisition program to acquire and store data in a spreadsheet file.

(3) Data analysis to carry out digital signal processing, calculate PWV and present results.

PWV Calculation

1. PWV calculation using peak detection. To calculate PWV using peaks, the location of the peaks must first be determined, so that the transit time of the wave between the peaks can be determined. It was found that the best method of peak

detection is the derivative of the curve method. If the first derivative of a curve is zero, then an extreme value can exist—either

a peak or a turning point. It is necessary to take the second derivative at this point—if this is also zero, then an extreme value exists. The second VI used to determine the PWV is PWVcalc, using the time separation between the located peaks.

2. PWV with pressure wave foot detection. The VI named PWV Foot determines the leading edge (foot) of the pressure wave at the upstream and downstream

locations. The VI named PWVcalc is again used to compute the PWV from the time separation between the two leading edges (‘foot-to-foot’ APWV or FFAPWV).

3. PWV with cross-correlation. The PWV determination with cross-correlation is done with the VI named CalcPWV. The VI is in two parts: a part for the initialization function and a part for the calculation of the CCAPWV.

In all cases PWV values are assembled in an array and the mean value, standard deviation and variance are calculated.

Sensor positioning

Sensor placement is critical to obtaining consistent measurements. A screw mechanism was first used to apply the sensors to the skin. But readings were very variable and so this technique was replaced by that in which the sensors are fixed to the skin by elastic strips. This led to better results.

Arm position is another critical feature of measurement. Two positions, normal and dependent, were analyzed in detail, using one test subject. In the normal position, the subject sits with the arm resting on a table. All test subject measurements were made in this position. In the dependent position, the subject sits with the arm hanging straight down.