 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to an apparatus and various methods used therein for
continuous non-invasive measurement of the pressure of a pulsatile fluid
flowing through a flexible tube over relatively long time periods, with
particular applicability to the measurement of human arterial blood
pressure and other related cardiovascular parameters.
2. Description of the Prior Art
Often a need arises to monitor pulsatile fluid pressure in a vessel where a
number of practical considerations preclude direct invasive measurement,
i.e. using an appropriate pressure sensor directly implanted through the
vessel and maintained in a suitable position within the fluidic flow.
Considerations of this nature include avoiding: contamination of the fluid
or its immediate environment by any foreign matter carried by the sensor,
coagulation of the fluid, corrosive deterioration caused to the sensor by
direct contact with the fluid, fluid loss from the vessel, or physical
damage to the flexible vessel that contains the fluid.
These considerations have particular applicability to the measurement of
arterial blood pressure of human (or other animal) patients or subjects.
In practice, invasive pressure monitoring generally entails a surgical
cut-down and arterial penetration using a hypodermic needle (cannula)
through which pulsatile fluidic forces attributable to arterial blood flow
are routed to a suitable pressure transducer. However, various medical
health care risks associated with invading the human body, such as
clotting, infection, emboli obstructions to flow, and/or major blood loss
generally limit the use of invasive blood pressure monitoring systems to
the most critical of acute care hospital patient monitoring situations. To
minimize these risks, invasive monitoring is almost always used in
conjunction with intravenous application of fluids to the patient.
Disadvantageously, the various fluidic, mechanical and electrical
components generally used in invasive systems are not only complex and
fragile, but also require specialized calibration before use, as well as
frequent surveillance by specially qualified staff. In spite of this
surveillance, invasive systems often do not remain calibrated during
prolonged periods of use and, as a result, often produce inaccurate and
potentially misleading patient blood pressure measurements.
Consequently, over the years several techniques have been developed for
non-invasive arterial blood pressure measurement. In general, these
techniques rely upon attaching an inflatable cuff around an extremity
(limb), which is typically an upper arm, of a human patient. Once
attached, the air pressure existing within the cuff is increased to a
value commonly referred to as "suprasystolic," i.e., a pressure in excess
of that minimally necessary to completely occlude a major artery existing
within the extremity and situated near its surface. Thereafter, blood
pressure is most commonly estimated by detecting well-known "Korotkoff"
sounds using a stethoscope, a microphone, or an ultrasonic detector placed
on the limb near the artery. These Korotkoff sounds are produced by the
artery and, more particularly, by disturbances in the arterial blood flow
due to partial occlusions of the artery caused by the externally applied
cuff pressure. As the cuff pressure decreases and the extent of occlusion
is reduced, various classic phases of sound change are usually heard until
the artery is no longer occluded by any appreciable amount. Specifically,
as the cuff pressure is reduced from suprasystolic, the maximum value of
pulsatile blood pressure commonly referred to as systolic pressure, is
usually taken to be equal to the cuff pressure at the time the first
Korotkoff sound is detected. Thereafter, the minimum or so-called
"diastolic pressure" value of pulsatile arterial pressure is usually
identified in conjunction with the occurrence of one of two other
Korotkoff phases: either the so-called fifth phase when silence occurs or
the so-called fourth phase which corresponds to a cuff pressure of about
5-10 mm(Hg) higher than that occurring at the fifth phase. Manual pressure
readings for systolic and diastolic are determined by identifying each
desired phase, and, as the cuff pressure continually decreases,
simultaneously noting the scale value in mm(Hg) that corresponds to the
height of a mercury column (or the pointer on an aneroid gauge) which is
pneumatically connected to the cuff air pressure. Devices of this sort are
commonly referred to as "sphygmomanometers."
Unfortunately, the accuracy of any non-invasive sphygmomanometer type blood
pressure measurement system, typified by that described above, is largely
dependent on the skill and hearing acuity of its user in detecting the
rather subtle sound changes (such as the very gradual transition to
silence after the fourth phase occurs), and, simultaneously therewith,
determining the exact level of the mercury column. In addition, dexterity,
sensory limitations and inexperience of the user; interference of
environmental noises, and the need to frequently calibrate the measurement
system often occur and all contribute to produce highly inconsistent
results. This inconsistency is a widely known characteristic of
sphygmomanometric systems.
Consequently, in an endeavour to minimize inconsistent results, many
attempts have occurred in the art to automate the process of
sphygmomanometric measurement. Specifically, these attempts involve using
electronic processing circuitry to automatically determine the desired
phases of Korotkoff sound change and the simultaneously occurring systolic
and diastolic cuff pressure values.
These attempts are typified by the systems disclosed in U.S. Pat. Nos.
3,581,734 (issued to Croslin et al on June 1, 1971); 4,245,648 (issued to
Trimmer et al on Jan. 20, 1981) and 4,271,844 (issued to Croslin on June
9, 1981). Each of these three patents discloses a computerized
sphygmomanometer measurement system in which an occlusive cuff is attached
around a limb of a patient. The cuff is then inflated, either manually or
automatically by an electrically driven air pump which is controlled
through either a computer or a hard-wired digital circuit. In each of
these systems, the cuff is inflated to a suprasystolic occluding pressure
prior to taking (sampling) any blood pressure measurement data. Then, by
automatically undertaking various detection and determination processes,
as well as deflating (bleeding-down) the occlusive cuff, these systems
attempt to eliminate many of the above-described manual steps that can
cause measurement error in manual sphygmomanometer systems known to the
art. Unfortunately, these automated sphygmomanometer systems are incapable
of reliably and consistently representing the true status of blood
pressure, in the same manner as provided by direct invasive monitors that
are widely accepted as the "standard of blood pressure measurement
accuracy".
Specifically, sphygmomanometer systems commonly produce misrepresentative
results due to a number of factors that are generally transparent to or
incapable of being compensated by the practioner-user. One such factor is
the impracticality of causing the pressure of any deflating occlusive cuff
known in the art to be made equal to the true peak systolic pressure value
of any one or more intra-arterial pressure waveforms such that the
measured cuff pressure is an accurate representation of systolic. This
impracticality results from the fact that any pulsatile intra-arterial
peak pressure value exists for only a short interval of time, (e.g.
usually less than 5% of the time). Inasmuch as the timing of cuff pressure
bleed-down is a random variable, the cuff pressure is typically lower than
true systolic peak pressure by random amounts, e.g., up to 10 mm(Hg),
depending on the deflation rate used before the desired Korotkoff or
pressure displacement waveform signal occurs--which indicates when the
cuff pressure is to be measured and designated as the systolic measurement
value. A second factor is the apparent lack of any uniform and accurate
diastolic determination method in systems known to the art. Specifically,
either one of two Korotkoff phases, i.e., the fourth and fifth phase, each
of which produces consistently different measurement values have found
wide use in prior art systems. Also, these diastolic measurement methods
known in the art often determine diastolic pressure as the value of
occlusive cuff pressure whenever it exceeds a certain threshold. Such a
threshold value is primarily dependent on the cuff pressure at the mean
arterial blood pressure instead of other more relevant and accurate
independent physiologic variables. Moreover, these methods, are generally
premised on an assumed linear relationship existing between amplitude
values at mean and diastolic pressure and linearly extrapolate the
diastolic pressure value based upon the mean pressure value. Accordingly,
these measurement methods have the unfortunate effect of assuming somewhat
erroneously, that a single fixed linear elasticity relationship defines
the stress/strain (e.g. pressure/displacement) characteristics of the
artery walls of all patients for whom blood pressure is to be
non-invasively measured--thereby resulting in an inaccurate determination
of the true diastolic pressure value. Lastly, a third factor is that
systolic and diastolic pressures commonly vary by differing amounts from
one heart-beat to the next due to several physiologic factors for both
normal and critically ill patients. Unfortunately, any combination of
these factors serves to over- or under-state not only the value of blood
pressure, but also more importantly changes in arterial blood pressure
occurring over time between successive measurements taken from any one
patient.
Moreover, these prior art systems not only lack the capability of
accurately portraying the arterial blood pressure associated with
individual heart-beats, but also disadvantageously they generally produce
only one systolic and diastolic reading during a measurement cycle that
can span between 20 and 100 successive heartbeats. To properly represent
the true status of blood pressure on a heart-beat to heart-beat basis, a
much higher sampling rate is necessary. However, if any of these
sphygmomanometer systems were used to measure arterial blood pressure
variations on a continuous heartbeat-by-heartbeat basis, then the
occlusive cuff would need to be repetitively and successively inflated to
a suprasystolic pressure and possibly to atmospheric pressure over many
short successive intervals of time, such as, for example, 10 times per
second, as is disclosed in U.S. Pat. No. 4,343,314 (issued to Sramek on
Aug. 10, 1982). Unfortunately, prevailing medical opinion is that any
patient wearing an occlusive cuff cannot be continuously subjected to
either elevated cuff pressures more than about 30% of the time during
which the cuff is being worn or repetitive cycling of cuff pressure
between suprasystolic and sub-diastolic pressures on the order of more
than once every one to three minutes, without experiencing significant
discomfort, trauma, and possible physiologic damage.
A typical repetitive cycling sphygmomanometer measurement system known to
the art which attempts to minimize patient discomfort is disclosed in U.S.
Pat. No. 4,378,807 (issued to Peterson et al on Apr. 5, 1983). As
described therein, a control circuit automatically initiates one cycle of
occlusive cuff inflation and bleed-down deflation only after a rather long
pre-defined interval of time typically on the order of 7.5 to 60 minutes
has elapsed. Unfortunately, pressure readings (one systolic and one
diastolic) are only taken at the conclusion of this relatively long
interval. As a result, the amount of measurement data is insufficient to
determine short and long term trends and variability in blood pressure, as
well as to identify, with any degree of reliability, any irregular
heart-beat pulsations. Thus, such a system is unsuitable for prolonged
continuous blood pressure monitoring of the critical patients.
Consequently, invasive pressure monitoring systems--even in spite of their
attendant health risks, as discussed above--are used to continuously
measure and display the pressure waveform for each heart-beat and to
compute the systolic, diastolic and mean pressure values based upon
averages of a number (typically 4-6) of successive heartbeats.
An alternate well-known scheme of non-invasive monitoring involves
measuring arterial wall displacement (i.e., radial distension of the
artery wall) produced by pulsatile arterial blood pressure and then
translating the measured displacement into an instantaneous blood pressure
value. These measurements and translations would, if performed at a
sufficiently rapid rate, appear to be continuous, i.e. result in the
display of an uninterrupted trend of sequentially-occurring pressure
waveforms showing substantial detail of the pulsatile nature of each
waveform, much in the same fashion as obtained through an invasive
monitor. See, for example, P. Flaud et al, "Pulsed Flows in Viscoelastic
Pipes. Application to Blood Circulation", Journal of Physics (France) Vol
35, No. 11, Nov. 1974, pages 869-882 and P. Flaud et al, "An Experimental
Device for Modelling Arterial Blood Flow," Review of Physical Applications
(France) Vol. 10, No. 2, March 1975, pages 61-67, which disclose that
radial displacement of the arterial wall is related to intra-arterial
blood pressure changes. However, the relative magnitude of wall
displacement is also directly related to the elasticity of the wall of the
arterial vessel. Unfortunately, arterial elasticity not only varies
significantly from patient to patient but also varies at different
locations along each artery, as well as at different times for the same
patient. Thus, a noninvasive pressure monitoring system that relies on
relative arterial wall displacement, requires that its measurements first
be calibrated against pressure measurements taken by a separate reference
device, such as an occlusive cuff, which would then serve as a calibration
reference for subsequent pressure values based upon arterial wall
displacement measurements.
Hence, in view of the drawbacks associated with prior art non-invasive
measurement systems, continuous blood pressure monitoring systems known
and used in the art are generally invasive and thus rely on intruding a
major artery of the patient.
SUMMARY OF THE INVENTION
An object of the present invention is to provide a system which can
continuously and non-invasively monitor the pressure of a pulsatile fluid
flowing through a flexible tube.
A specific object is to provide an accurate continuous non-invasive blood
pressure monitoring system which can be attached to a patient and operated
for either relatively brief or substantially prolonged periods of time
without any physiologic risks and/or significant discomfort to that
patient.
Another specific object is to provide such a monitoring system which
generates a substantially continuous record of pulsatile blood pressure
activity and associated numeric measurement parameters.
Another specific object is to provide such a system in which its accuracy
is substantially unaffected by the skill of the user, and for which
minimal retraining is required for persons already experienced in the use
of available pressure measurement instruments and techniques.
Another specific object is to provide such a system which deflates a blood
pressure cuff, through one or more controlled substantially linear rates.
Another specific object is to provide such a system which compensates the
systolic and diastolic measurements for the affects of hemodynamic
variability thereby producing measurement values that are consistent with
those produced by invasive monitoring systems.
Another specific object is to provide such a system which is entirely
self-contained.
Another specific object is to provide such a system which automatically
adjusts for a multitude of different, typically non-linear factors such as
variations in arterial elasticity and the types and relative amounts of
intervening tissue existent between a patient's artery and a non-occluding
pressure sensing cuff.
Another specific object is to provide such a system, which to ensure
accurate consistent measurements, automatically determines whenever,
during a period of continuous monitoring, it requires re-calibration and
then effectuates any such re-calibration(s).
Another object is to provide a calibration process which can be used as a
basis for calibrated measurement of pulsatile arterial activity, such as,
for example, systolic, diastolic, and mean blood pressure parameters using
many known arterial measurement systems, for durations in excess of a few
seconds, and which are typified by non-invasive electrical impedance and
strain gauge plethysmography or invasive perivascular sensing methods
(which detect blood flow, volume, or velocity as a function of various
physiologic parameters that may be taken to be proportional to measured
variations in intra-arterial pressure).
Another specific object is to provide such a calibration process that is
compatible with other well-known types of sphygmomanometric systems that
utilize an occlusive cuff, including those employing microphones and
ultrasound flutter principles which detect arterial phenomena as
measurement signals during semi-occluded blood flow conditions.
Lastly, another specific object is to provide such a system which utilizes
calibration processes of the occlusive cuff to optionally measure discrete
systolic and diastolic pressures on an intermittent basis at pre-set time
intervals in a similar fashion to other automatic non-invasive pressure
measurement products known in the art.
These and other objects are accomplished in accordance with the teachings
of this invention by first undertaking a calibration phase (procedure)
comprised of: determining the blood pressure occurring in relation to
various initial conditions of arterial blood flow and ascertaining the
values of a plurality of coefficients, each of which is associated with a
corresponding term in a pre-defined function that characterizes blood
pressure values in relation to arterial wall displacement; and second,
undertaking a continuous monitoring phase comprised of: continually
measuring subsequently occurring arterial wall pressure displacement
waveform values, ascertaining each subsequently occurring blood pressure
value as the pre-defined function of each corresponding measured arterial
pressure displacement waveform value, and automatically re-calibrating the
system to the patient after the expiration of a pre-defined but adaptively
changeable time or in response to the occurrence of any one of a plurality
of pre-defined events. This interval, i.e. the time between the
occurrences of successive re-calibrations, is adaptively changed in
accordance with the results of at least one prior re-calibration and/or
whenever significant changes in the trend of arterial wall displacement
values occurs.
In accordance with the specific embodiment disclosed herein, two separate
inflatable cuffs (a relatively high pressure occlusive cuff and a
relatively low pressure waveform sensing cuff) are affixed to different
locations proximately situated to major arteries of one or two limbs of a
patient's body. A computer in conjunction with various pneumatic
components effectuates the process of inflation (pressurization) and
deflation of each cuff as well as the data acquisition from each.
Instantaneous arterial blood pressure is characterized in terms of a
parabolic function of arterial wall displacement: namely f(x)=ax.sup.2
+bx, where x is proportional to changes in arterial wall displacement
measured from a constant reference value, and (a) and (b) are coefficients
having values that primarily depend upon various physical characteristics
(such as the elasticity of the arterial wall and interspersed biological
tissue--existing at the site where blood pressure is being measured, and
the cuff material itself) and various blood pressure values determined
during the calibration phase through use of the occlusive cuff.
Operation of the blood pressure measurement system occurs, via essentially
a two-phase approach. During the first, or "calibration", phase, the
computer automatically inflates (pressurizes) the occlusive cuff to a
pre-defined suprasytolic value, typically on the order of 150 mm(Hg), and
also inflates the waveform sensing cuff to a relatively low pressure of
approximately 40 mm(Hg). During this time, the computer automatically
checks the integrity of both cuffs to determine whether any significant
pneumatic leakage exists anywhere in the system and confirms that both
cuffs are properly affixed to the patient. Once both cuffs have been
inflated, the computer causes the pneumatic components to bleed down the
pressure in the occlusive cuff at a controlled rate while maintaining the
pressure of the waveform sensing cuff constant at a value of approximately
40 mm(Hg). During this controlled bleed-down, arterial pressure
displacement waveform information is sensed through instantaneous pressure
variations (perturbations) occurring in both the occlusive and the
waveform sensing cuffs.
The resulting displacement waveform information from both cuffs is
digitized and resulting sample values are stored by the computer. These
samples are then processed, via several different techniques, to determine
systolic and diastolic occlusive cuff pressures, as well as the values of
the coefficients (a) and (b).
The particular methods for determing systolic and diastolic occlusive cuff
pressure values adjust for heart-beat to heart-beat variability, random
bleed rate errors, and patient movement artifacts. Specifically, both the
systolic and diastolic pressure determinations are dependent upon analysis
and weighted calculations of two groups of at least four pressure
displacement waveforms which generally occur when the occlusive cuff
pressure is at or near true systolic pressure, first, and then true
diastolic pressure, second, during bleed-down. Simultaneously therewith,
measurements of the same two groups of pressure waveforms are made using
the constant low pressure waveform sensor cuff. Simultaneously with (or
shortly after--in the event both cuffs are positioned on one limb rather
than on opposite limbs) the determination of systolic and diastolic
pressure values through occlusive cuff measurements, peak and trough
values, associated with the sequence of waveforms measured from the
waveform sensor cuff, and averages (denoted as base level values) of these
peak and trough values are computed. Shortly thereafter, the occlusive
cuff pressure is exhausted to ambient (i.e. atmospheric), and the peak and
trough values, the corresponding base level values as well as the
corresponding systolic and diastolic occlusive pressure values are all
used to determine coefficient values (a) and (b) contained in the arterial
wall displacement/pressure function. After these coefficients are
determined, the computer uses the displacement/pressure function to
calculate a set of values for subsequent storage in a look-up table. This
table consists of the blood pressure values that correspond to a
relatively large number of uniformly-spaced arterial wall displacement
values which span an entire pre-defined numeric range.
The second or "continuous" monitoring phase, based on the newly calculated
look-up table, commences immediately upon conclusion of the calibration
phase. During this second phase, data in the look-up table is used to
convert actual arterial pressure displacement waveform sample values into
blood pressure waveforms for both display and subsequent calculation of
various numeric measurement parameters. Specifically, indivi | | |
