 |
Description  |
|
|
BACKGROUND AND SUMMARY OF THE INVENTION
In general, previous instruments utilizing photoelectric techniques have
been limited to monitoring blood flow and have not provided accurate blood
flow rate measurements. Some of these previous techniques for blood flow
rate monitoring have used a plethysmograph to monitor blood flow through
an organ, typically the pinna of an ear or one of the patient's fingers.
See, for example, an article in the American Journal of Physiology, 1940,
Volume 130, No. 1 by Alrick B. Hertzman and John B. Dillon, entitled
"Distinction Between Arterial, Venous and Flow Components in Photoelectric
Plethysmography in Man." Also see, for example, U.S. Pat. No. 3,796,213
issued to Frederick Richard Neason Stephens on Mar. 12, 1974 and entitled,
"Perfusion Monitor."
Photoelectric techniques for monitoring or measuring blood perfusion are
based on the phenomena that changing blood volume gives rise to a changing
light absorption and hence changes in the amplitude of a signal produced
by a photosensitive device. Photoelectric plethysmographs are used
extensively for studying relative changes in skin blood flow and and as a
transducer for heartrate monitors. However, problems associated with
calibration, linearity, and stability, have all but eliminated the use of
photoelectric plethysmography for accurate non-invasive blood flow
measurement.
Also, oximeters have been used to measure the quantity of oxygen in a
patient's blood. However, in many instances the usefulness of the oximeter
has been limited because certain patients, e.g., those in post-surgical
recovery, have sufficient peripheral vaso-constriction to limit the
application of the oximeter. This is because while the oximeter can
indicate when the quantity of blood in a tissue being measured is
sufficient it cannot give an indication of its flow state. Consequently,
because of the insufficient blood flow, the oximeter typically analyzes a
mixture of venous and arterial blood and thus gives an A.sub.os
measurement of smaller value than it should be. It is therefore necessary
to accurately measure the blood flow rate through the tissue to calibrate
the oximeter readings, but, as stated above, prior art perfusion meters
have not provided the desired accuracy and convenience. Hence, to
determine the flow rate of blood within a tissue area under test and to
provide a method for estimating the accuracy of an oximeter reading, the
perfusion meter in accordance with the present invention has been
developed.
The physiological parameter that makes possible all photoelectric
plethysmographic techniques is the pulsatile color changes associated with
blood flow through the microcirculatory vessels. As mentioned above,
photoelectric plethysmographic techniques have been used for studying
relative changes in skin blood flow but many problems are encountered when
attempts are made to make accurate perfusion measurements. For example, a
primary cause of linearity problems is the response characteristics of the
photosensitive device used in the plethysmograph.
All known photosensitive devices, with the exception of a photodiode which
has a low level output, have non-linear response characteristics similar
to that shown in FIG. 1. Also photosensitive devices have some variation
with parameters other than light but in general, the more sensitive the
device the worse the stability problem. For example, photoconductive cells
typically suffer from large variations with temperature and light history
problems so that their usefulness has definitely been limited, at least as
it applies to accurate blood flow rate measurement. Phototransistors,
while being fairly stable, suffer from non-linearity problems as well as
being susceptible to radio frequency interference because of their
rectifying junctions. Hence, even with calibration, which was only
temporary at the best, prior art photoelectric plethysmographic techniques
(transmissive and reflective) could not reliably be used for accurate
non-invasive blood flow rate measurement.
In accordance with the preferred embodiment of the present invention, a
light emissive device provides illumination through body tissue to a
photosensitive device. A signal is produced which represents the log of
the transmitted light. Changes in the log of the signal produced in
response to the transmitted light are utilized to produce indications
representing changes in the pulsatile blood flow rate.
DESCRIPTION OF THE DRAWINGS
FIG. 1 is a graph showing the characteristic response of a typical
photosensitive device to illumination.
FIG. 2 illustrates the structure of a typical capillary bed.
FIG. 3 shows a block diagram of the perfusion meter of the preferred
embodiment.
FIG. 4 shows the waveform of the signal obtained from the Averager circuit
of the preferred embodiment.
FIG. 5 is an illustration of the portable case and perfusion index display
of the preferred embodiment.
FIGS. 6A, 6B, 6C and 6D comprise an illustration of the optical
plethysmograph of the present invention.
FIGS. 7A and 7B, taken together, are a schematic diagram of the circuitry
of the preferred embodiment.
FIG. 8 is a simplified schematic diagram of the operation of the input and
logging circuitry of the preferred embodiment.
FIG. 9 shows the matching of the light emissive device and the
photoconductive device of the preferred embodiment.
FIG. 10 is a graph showing the absorption spectra of blood through a
typical ear at different oxygen concentrations.
FIG. 11 is a graph of the response of the preferred embodiment to varying
tissue samples.
DESCRIPTION OF THE PREFERRED EMBODIMENT
It is in the capillaries that the most purposeful function of circulation
occurs, namely the interchange of nutrients and cellular excreta between
the tissues and the circulating blood. To accomplish this, an estimated
3,600,000,000 capillaries are distributed throughout the human body so
that they are rarely more than 20 microns away from a single functional
cell.
FIG. 2 illustrates the structure of a "unit" capillary bed and shows the
arterial blood entering via an arteriole 10 passing through a capillary
bed and leaving by way of a venule 15. Blood from the arteriole 10 passes
into a series of meta-arterioles 20 and from there into capillaries 25.
Some capillaries are relatively large and are called the preferential
channels, while the smaller capillaries are sometimes referred to as true
capillaries. After passing through the capillaries or through an A-V
shunt, the blood passed through a venule and then into the veins. The
veins are capable of constricting and enlarging and thus they can store or
pump the blood back into general circulation.
The arterioles 10 are highly muscular and can change their diameters many
times. The meta-arterioles 20 do not have a continuous muscular wall, but
smooth muscle fibers encircle the vessels at intermediate points. Smooth
muscle fibers, called the precapillary sphincters, usually encircle the
capillaries at their inlet. The venules 15 are considerably larger than
arterioles 10 and have a much weaker muscular coat.
Vasomotion, i.e., the intermittant or discontinuous flow of blood through
the capillaries, is caused by the intermittent contraction and relaxation
of the meta-arterioles and precapillary sphincters. Vasomotion regulation
is primarily a function of the concentration of oxygen in the tissues.
Typically, only 10-15% of the tissue capillaries are open.
Despite the fact that blood flow through each capillary is intermittant,
there are so many capillaries in a given area of body tissue that their
overall function becomes averaged, i.e., there is an average rate of blood
flow through each tissue capillary bed. Likewise, the tissue perfusion
measured by the preferred embodiment is the average blood flow through the
unit tissue area.
Although each capillary has a minute cross-sectional area, there are so
many capillaries in the circulatory system that the total cross-sectional
area is very large. Since the velocity of blood flow varies inversely with
the cross-sectional area, it can readily be appreciated why the blood
flows so slowly in the capillaries. This combination, blood velocity and
blood crosssectional area, gives vascular color to the skin. The hue is
determined by the rate at which the blood flows through the capillaries,
while the intensity of the color is determined by the thickness of the
blood layer. Hence, both the diameter of the microcirculatory vessels and
the number of conducting vessels influence color brilliance, and the
transmissivity of a given tissue area.
These pulsatile color changes associated with blood flow through the
microcirculatory vessels are the physiological parameter upon which all
photoelectric plethysmographic techniques are based. As stated above,
photoelectric plethysmographs have been used for studying relative changes
in skin blood flow, but their linearity, stability, and calibration
problems have all but eliminated their use for non-invasive perfusion
measurements. However, using the techniques of the preferred embodiment,
accurate non-invasive blood flow rate measurement is now possible.
The perfusion meter of the preferred embodiment is a single-wave length
device based on a Beer's law model of the optically absorbing material in
the ear or other tissue. In this model, it is assumed that, in the absence
of any field motion, the only short-term change in the absorbent spectrum
is due to the pulsatile color changes associated with blood flow through
the microcirculatory vessels.
To understand this model, consider a single light absorber of thickness d
concentration c which receives light I.sub.o and transmits light I. The
transmittance of the light absorbing substance at wave length .lambda. is
given by
T(.lambda.) = I/I.sub.o = e .sup.-E(.lambda.)cd = E.sup.-A(.lambda.)
and the absorbance A(.lambda.), of the substance at wave length .lambda. is
given by
A(.lambda.) = -1n(I/I.sub.o) = E(.lambda.)cd
In both of the above equations, E(.lambda.), is a proportionality constant
known as the molar extinction coefficient which varies as a function of
the material and wave length of light.
Note that if a mixture of two or more substances, e.g., a first substance
having a concentration C.sub.1 and a thickness D.sub.1, a second substance
having a concentration C.sub.2 and a thickness D.sub.2, . . . , and an Nth
substance having a concentration C.sub.N and a thickness D.sub.N, are
contained in one sample, the absorbances of each will be added as follows:
A(.lambda.) = -1n(I/I.sub.o) = E.sub.1 (.lambda.)C.sub.1 d.sub.1 +E.sub.2
(.lambda.)C.sub.2 d.sub.2 +. . . +E.sub.N (.lambda.)C.sub.N d.sub.N
if this happens to be a tissue sample with pulsatile blood flow, one of the
above absorbent spectra, (HbO.sub.2 + Hb), will be changing with time,
while in the absence of field motion, all others should remain very nearly
constant. Thus, by taking the derivative of the above equation with
respect to time, the following is obtained:
##EQU1##
Furthermore, to relate linearity to pulsatile blood flow the following
mathematical manipulation can be done.
.DELTA.1n(I/I.sub.o) = .DELTA.1n I - .DELTA.1n I.sub.o = -.DELTA.E.sub.b
(.lambda.)C.sub.b d.sub.b
.DELTA.1n I = -.DELTA.E.sub.b (.lambda.)C.sub.b d.sub.b
This resulting equation, a time-varying log function, indicates that all
fixed absorbers and system gain factors can be ignored and that in the
absence of field motion, the changes in the log of the transmitted light
are directly proportional to the changes in the blood thickness --
concentration product, that is, the changing blood volume in the field of
view.
Deoxyhemoglobin and oxyhemoglobin are the main absorbers of the blood and
they have different absorption spectra, as is illustrated in FIG. 10.
Choosing a wavelength between 6200A and 8000A gives greater sensitivity to
oxyhemoglobin while a wavelength between 8000A and 10000A gives slightly
greater sensitivity to deoxyhemoglobin. The ideal wavelength is
approximately 8000A for there both deoxyhemoglobin and oxyhemoglobin have
the same absorption. Unfortunately, neither a source nor a detector is
commercially available at 8000A so a compromise solution must be chosen.
It is desirable to match the spectral response of the light emissive device
to the photodetecting device to minimize ambient light interference and
system power requirements. The former of these is important in order to
give the best possible signal to noise ratio. The latter is particularly
important if the system is battery operated. FIG. 9 shows that with proper
selection of a CdSe photoconductive cell and a GaP red solid state lamp or
light emitting diode (LED) an extremely well matched spectral response may
be obtained. The combination of these two devices yields a sevenfold
improvement in current drain over that obtained with the use of an
incandescent lamp. Additionally, the LED provides for a safer (less heat)
system as well as for greater reliability. As for speed of response, the
selected CdSe photoconductive cell is fast enough to handle the relatively
slow cardiac pulse activity.
With this selection complete, it is now essential that the historical
problems associated with calibration, linearity and stability be
eliminated. This is accomplished by using standard components having a
sensitivity response as shown in FIG. 1, and coupling them with an element
having a logarithmic response. Any non-linearity in a
resistance-illumination curve is acceptable as long as it approximates a
straight line when plotted on log-log paper.
A graph of responses obtained using the preferred embodiment is shown in
FIG. 11.
The relationship between illumination and photoconductive resistance is
given by an equation of the form:
R = AI.sup.k
where R = photoconductive resistance
I = illumination
A and k are constants and are dependent upon material
To fully capitalized on the benefits of the logging operation, a unique
probe had to be developed which would work equally well on fingers and
toes as well as on the pinna of an ear. Historically, a transillumination
has been used on ears whereas reflectance has been used on fingers and
toes. The latter technique has shown considerable variability associated
with application pressure and thus its usefulness, even with the logging
operation, would be questioned. FIGS. 6A-D shows an optical probe that
allows for equal usefulness on an ear, or finger and can very easily be
applied to either simply by squeezing the sides of the "clamshell" as
shown in FIG. 6C. In all cases, the application pressure, applied to
either ear or finger, is not sufficient to squeeze the blood vessels and
change the measurement.
Although the slope of the response characteristic in FIG. 11 is not
perfectly constant, its deviation is small. That is, the changes in slope
-- from a thin white ear to a thick black ear to the tip of a finger --
are small. Similarly, light history and temperature drift simply shift the
operating point along the constant slope line. Thus, regardless of light
history, temperature or sample transmission, the constants of the equation
relating photoconductive resistance to illumination are changing but by a
small percentage. This then, shows that logging provides for linearity,
eliminates temperature and light history stability problems, and allows
for reliable perfusion measurements on any tissue regardless of color or
thickness. FIG. 8 shows a light emissive device emitting illumination,
I.sub.o, through tissue having pulsatile blood flow. If a constant
voltage, V', is applied to the photoconductive resistance, R = AI.sup.k,
the current, i, is given by: i = V'/R. Logging this current yields:
1n i = 1n V' - 1n R
= 1n V' - 1n (AI.sup.k)
= 1n V' - 1n A - k 1n I
the time varying components of this equation are given by:
##EQU2##
or:
.DELTA.1n i = -k .DELTA.1n I
earlier, however, it was shown, in the absence of field motion, the changes
in the log of the transmitted light to be directly proportional to the
changes in the blood thickness-concentration product. Therefore:
.DELTA.1n i = k.DELTA.E.sub.b (.lambda.)D.sub.b d.sub.b
This equation shows that the only system determined variable is, k, which
is a function of the slope of the photoconductive cell. By spectifying
limits on the slope, instrument to instrument differences may be properly
controlled. As the slope changes but by a small percentage under various
illuminations, the historical problems of linearity and stability are
solved.
Because a pn semiconductor junction exhibits a repeatable logarithmic
relationship between current and voltage, it can be exploited to perform
the needed logging operation. Since the diode connected transistor
conducts the current from the photoconductive cell, and as the voltage
drop across that junction approximates the log of that current, the
changes in this voltage correspond to the changes in blood volume.
In FIG. 3 there is shown a block diagram of the perfusion indicator in
accordance with the preferred embodiment. A light source 305 is driven by
a current source 310. The emitted light, Io, produced by light source 305
is directed towards tissue 300; the transmitted light, I, falls on
photo-responsive device 335. A signal on a line 336 is produced by photo
responsive device 335 in response to said transmitted light.
A logging circuit 340 produces an output signal on a line 341 which bears a
logarithmic relationship to the signal received on line 336. The signal on
line 341 is capacitively coupled to a bandpass amplifier 350. The passband
of amplifier 350 is approximately 0.5-11 Hz, this being the frequency band
of interest with respect to blood flow.
An amplified pulse signal is produced on line 351 and coupled to a clamping
circuit 360 which clamps the most negative portion of the pulse signal at
zero volts. This produces a ground referenced waveform on line 361 which
is then averaged by an averager circuit 370. The averager output signal on
a line 371 has a DC voltage component (time integration of volume pulse)
that is proportional to the average blood flow, effectively the volume
pulse amplitude-heart rate product, and an AC voltage component that is
proportional to the pulse amplitude. This waveform is shown in FIG. 4.
The signal on line 371 is input to analog to digital converter 380 which
simultaneously produces a 7 level analog staircase voltage and decoding
for a 7 bit digital response. The average waveform is constantly compared
with the staircase voltage and if the comparison is met, a corresponding
element in the perfusion index display is lit. Thus, a simple, inexpensive
A to D conversion with built-in display is provided by this circuit.
The seven elements of the perfusion indicator display 390 are constantly
strobed producing a bar graph type of display in response to the signal on
a line 371. This bar graph display provides an indication called the
perfusion index which is a measure of the pulsatile blood flow within the
body tissue. Also, as mentioned above, this reading can be interpreted to
indicate the possible error in an oximeter reading.
Referring to FIG. 5, there are shown indicators 510 which are energized in
response to the blood flow rate within the tissue under test. These
indications may be interpreted per legend 520 as indicating the probable
oximeter error or correction factor. These indications may also be
interpreted using legend 530 as a direct measurement of the perfusion,
i.e., the blood flow rate.
The representation provided by indicators 510 may be interpreted as a bar
graph display. The illuminated length of the bar graph, starting at the
leftmost indicator and continuing to the rightmost illuminated indicator,
represents the average blood flow rate. The pulse rate is indicated by the
flashing of the rightmost indicators. The strength of the pulse is
indicated by the number of flashing indicators; more indicators flash on
and off in sequence as the pulse strength increases.
Measurements may be made after stimulating the ear or other body tissue to
produce the maximum blood flow. This is done by brisk rubbing and/or
gentle heat. The plethysmograph is then placed over the pinna of the ear
or other body surface and the interconnecting cord between the
plethysmograph and the portable case 510 is draped to minimize
interference from motion artifacts. A key 540 in the center of the
indicator case is depressed and held. In less than ten seconds the bar
graph display 510 settles to a reading which is a measure of the pulsatile
blood flow rate within the ear or other tissue.
In FIG. 7 there is shown a schematic diagram of the circuitry of the
preferred embodiment. A battery 320 is coupled through a power switch 716
to a power supply regulator circuit 720 which supplies 1-20 milliamps of
current at 4.5 volts on line 723. Whenever the power is on, low voltage
comparator circuit 325 continuously monitors the battery voltage 320 and
automatically turns off the perfusion index display 390 and illuminates
low battery voltage indicator 330 if the battery voltage is below a preset
limit.
A current source is provided by circuit 720 and resistor 310. The current
provided on a line 716 is used to drive the LED light source 701 within
ear probe 710. Current is supplied to a photoconductive device 702 within
an ear probe 710 by logging circuit 730.
Transistor 731 is connected to operate as a diode. The logarithmic
properties of the PN semiconductor junction, as mentioned above, between
current and voltage is exploited to perform the logging function. As shown
earlier, the changes in this junction voltage correspond to the changes in
blood volume. Inductor L1 and capacitor C10 form and L-C high frequency
filter which helps eliminate unwanted interference. Capacitors C8 and C9
further reduce any high frequency interference.
The output from logger circuit 730 is capacitively coupled via capacitor
733, to low pass amplifier circuit 350. The combination of amplifiers
within this circuit forms the 0.5-11 Hz bandpass amplifier and has a net
gain of about 3,000, i.e., 70 DB, within the passband.
The output from amplifier circuit 350 is capacitively coupled by capacitor
750 to averager circuit 370. Circuit 370 acts as an averager while circuit
360 is the clamping circuit. The clamping function is performed by
amplifier 760 which has negative feedback through diode 761 to prevent the
inputs of amplifier 760 and 763 from ever going negative. Thus, the entire
pulse waveform is positive with reference to ground except for the input
offset voltage of amplifier 760.
This ground referenced waveform on line 766 is then averaged by circuit
370. The output signal on line 371 has an amplified average, i.e., a DC
voltage proportional to the average blood flow, and a non-amplified pulse
waveform, i.e., an AC voltage waveform proportional to pulse strength.
Oscillator 780, shown in FIG. 7B provides the signal on a line 781 to
decade counter and decoder 790 within analog to digital converter circuit
380. Whenever the reset line, pin 15, of decade counter and decoder 790 is
low, the decade counter is counting and one of the ten outputs will be
decoded and an output signal therefrom will remain high for one clock
period. Seven of these output signals are used to generate a staircase
voltage on a line 791. Also, these seven outputs from decade counter and
decoder 790 are used to strobe the LED display devices 797. Note that
whenever the battery voltage is sufficiently low to trigger low voltage
comparator circuits 325, the enable signal on line 789 is removed and thus
staircase voltage generator 778 and indicator displays 797 are disabled.
Resistors 771, 772, 773, 774, 775, 776, and 777 are chosen to give a
staircase voltage at the input terminal of amplifier 778. The output on
line 792 from amplifier 778 is a staircase waveform comprising seven
voltage values varying from 0 volts DC to 3.6 volts DC in 0.6 volt
increments. Each of the voltages appears for one full clock cycle and
reoccurs every ten clock cycles. Amplifier 779 completes the analog to
digital conversion by continuously comparing the output from averager
circuit 370 with the various steps of voltage.
If the output signal on line 371 exceeds the staircase voltage, the output
on line 792 from amplifier 779 goes high and activates the current source
800 for the perfusion indicator display 797. Thus, whenever this condition
is met a current source 800 is activated and the appropriate buffer gate,
i.e., gates 782, 783, 784, 785, 786 and 787, are turned on and current is
passed through one of the LED indicators in display 797. Note that
indicator display 797 will provide two separate pieces of information. One
is that the number of LEDs which are lit in display 797 will correspond to
the DC level of the signal output by averagers circuit 370. Also, the
number of indicators in display 797 which are flashing on and off will
indicate the strength of the pulse through the tissue under test, the more
lights that are flashing on and off the stronger the pulse rate.
* * * * *
|
|
 |
|
 |
Description |
|
