 |
Description  |
|
|
BACKGROUND OF THE INVENTION
The degree of oxygen saturation of arterial blood is often a vital index of
the condition of a patient. Whereas apparatus is available for making
accurate measurements on a sample of blood in a cuvette, it is not always
possible or desirable to withdraw blood from a patient, and it is
obviously impracticable to do so when continuous monitoring is required.
Therefore, much effort has been expended in devising an instrument for
making the measurement by non-invasive means.
One approach has been to substitute a portion of the body such as a lobe of
the ear for the cuvette and measure the difference in light absorption
between oxyhemoglobin HbO.sub.2 and deoxyhemoglobin Hb. Unfortunately,
however, complications are introduced by the presence of a large number of
light absorbers other than HbO.sub.2 and Hb, including, for example, skin
and hair.
The following considerations are fundamental to the problem. As blood is
pulsed through the lungs by heart action a certain percentage of the
deoxyhemoglobin, Hb, picks up oxygen so as to become oxyhemoglobin,
HbO.sub.2. By medical definition the oxygen saturation SO.sub.2 =HbO.sub.2
/(Hb+HbO.sub.2). It is this fraction which is to be determined. From the
lungs the blood passes through the arterial system until it reaches the
capillaries at which point a portion of the HbO.sub.2 gives up its oxygen
to support the life process in adjacent cells. At the same time the blood
absorbs waste matter from the cells and is made to flow steadily back to
the heart by the vascular system.
Increasing the blood flow or perfusion in the ear increases the amount of
blood and therefore the amount of HbO.sub.2 reaching the capillaries to
such degree that the oxygen thus supplied can far exceed the amount of
oxygen consumed by the cells, thereby making the oxygen content of the
venous blood nearly equal to the oxygen content of the arterial blood.
Inasmuch as the light must obviously pass through both arterial as well as
venous blood, the measurement of the relative amounts of Hb and HbO.sub.2
yields an oxygen saturation measurement of SO.sub.2 that approaches the
value for arterial blood alone. Unfortunately, the condition of some
patients, for example a patient in shock, is often such as to prevent the
attainment of sufficient perfusion to yield highly accurate results.
Wood, as described in his U.S. Pat. No. 2,706,927, suggested a way of
computing the oxygen saturation from measurements of light absorption at
two wavelengths taken under two conditions, (1) a "bloodless" condition in
which as much blood as possible is squeezed from the earlobe and (2) a
condition of normal blood flow. It was hoped that the "bloodless"
measurement would be affected only by the absorbers other than blood and
that the normal blood flow measurement would be affected by both the other
absorbers and the blood so that a comparison of the readings would
indicate the absorption by the blood alone. Unfortunately, the accuracy of
the measurements is seriously impaired, not only by the fact that
squeezing does not eliminate all the blood but also because it changes the
optical coupling between the ear and the optical apparatus. Furthermore,
because of wide variations between patients in the effect of absorbers,
such as the pigment of the skin and its thickness, a separate calibration
must be made for each patient and for each measurement.
Many of these problems have been overcome by apparatus suggested by Shaw in
his U.S. Pat. No. 3,638,640 in which light absorption measurements are
taken at a number of wavelengths of light. However, in this as well as in
all other prior art apparatus, good results have depended on increasing
the perfusion in the member of the body being measured so that the blood
therein is as close to arterial blood as possible. Whereas perfusion can
be increased by artificial methods to the point where accurate results are
obtained, there are many situations when the patient's condition makes
such methods undesirable or even impossible.
BRIEF DESCRIPTION OF AN EMBODIMENT OF THIS INVENTION
If Hb and HbO.sub.2 are the only light absorbers of significance in the
arterial blood stream, the degree of oxygen saturation can be accurately
determined, in accordance with this invention, in the following general
way. Light of one wavelength and light of another wavelength are
sequentially directed to a given area on a finger, a lobe of the ear, or
other body member. Photosensitive means are positioned so as to produce a
first electrical signal that is proportional to light of the one
wavelength after a portion of it has been absorbed by the body member and
a second electrical signal after a portion of light of the other
wavelength has been absorbed by the body member. When the heart forces
more blood into the arterial system, there is more volume of blood in the
body member so that light of both wavelengths is attenuated more than when
the heart is at rest. Thus, both first and second electrical signals have
maximum and minimum peaks occurring during a heart cycle. It is important
to note that the difference between the peaks is due entirely to the
pulsatile arterial blood flow and that it is completely unaffected by
light absorbers that attenuate the light by constant amounts throughout
the heart cycle.
For reasons that will be explained at a later point, the light absorbed by
any absorber is directly proportional to the log of the light after it has
been attenuated by the absorber. Means are provided for deriving a first
output signal that varies as the peak to peak amplitude of the log of said
first electrical signal, and means are also provided for deriving a second
output signal that varies as the peak to peak amplitude of the log of said
second electrical signal. Each ratio of the first and second output
signals corresponds to a different percent of the oxygen saturation
SO.sub.2. Therefore, the percent of SO.sub.2 can be determined from a
graph or it can be automatically derived from a read only memory that is
programmed with the percent SO.sub.2 that corresponds to each ratio. Of
course this requires that the ratio be digitized. It is also possible to
use a microprocessor to make the calculation.
In some cases it may be desired to also determine the percent saturation of
another light absorbing blood component such as carboxyhemoglobin, HbC. To
do this three wavelengths of light are separately and sequentially
directed to the member of the body selected. The photosensitive device and
the logging means are the same as before but the voltage at their output
is now sequentially applied to three channels in synchronism and phase
with the sequential operation of the lights. Each channel is the same as
before, but because there are three of them the percentage saturations of
oxygen and the other components cannot be determined by a simple ratio.
However, as will be explained, the pulsatile output signal of each channel
is equal to the sum of the absorptions of the three absorbers of interest
for the particular wavelength of light in that channel. This yields three
simultaneous equations that can be solved. It is most convenient to apply
the signals to an A.D. converter and then to a microprocessor. The
percentage saturation of any number of pulsatile absorbers can be
determined in this way simply by using the same number of different
wavelengths of light and the same number of channels. As will be shown,
apparatus constructed in accordance with this invention can readily yield
other valuable information such as cardiac output, perfusion index and
heart rate.
THE DRAWINGS
FIG. 1 is a diagram of an oximeter utilizing light of two wavelengths,
FIG. 2 includes graphs showing the way the molecular extinction coefficient
varies with the wavelength of light of Hb, HbO.sub.2, and HbC,
FIG. 3 includes graphs showing the relationship between the ratio of the
pulsatile signals provided by the two channels of the oximeter of FIG. 1
vs. the percentage oxygen saturation for combinations of the two, and
FIG. 4 is a diagram of an oximeter utilizing light of three wavelengths in
order to detect the percentage of a third light absorber in the arterial
blood.
DESCRIPTION OF AN EMBODIMENT OF THE INVENTION
In the apparatus illustrated in FIG. 1, the timing of all circuit functions
is determined by a sequence controller 2. At an output 4, it provides a
series of equally spaced pulses 4' at a pulse repetition frequence f. In
response to these pulses a multiplex drive 6 provides pulses 8' on an
output lead 8 and pulses 10' on an output lead 10. The positive pulses 8'
and 10' have a frequency of f/2, a duration equal to the time between
successive pulses 4' and are interleaved. Light emitting diode 12 that
emits light of a wavelength .lambda..sub.1 and light emitting diode 14
that emits light of a wavelength .lambda..sub.2 are connected between the
leads 8 and 10 in such polarity that the diode 12 is turned on during the
pulses 8' and the diode 14 is turned on during the interleaved pulses 10'.
Although the pulses 8' and 10' are shown as having the same amplitude, the
amplitudes may be different if this is required to cause the intensity of
the light emitted by each diode to be as desired.
One embodiment is an array of suitably placed light emitting diodes. This
array may be comprised of discrete light emitting diodes or diode chips
mounted within a common package. In either case, light from the diodes 12
and 14 is passed through an optical diffuser 15. If desired, light pipes
could be used to conduct light from each light emitting source to the
input of the diffuser 15. Light from either diode appears at the output
surface of the diffuser 15 so that it will pass into the same area of the
body member with which the output of the diffuser is in contact.
It is intended that a member of the body such as a finger, as shown, or the
lobe of an ear, be inserted in the path 16 between the diffusor 15 and a
collection lens 30 of a photosensitive device 32 which would normally be
in optical contact with the finger or other body member.
The current, i, generated by the photosensitive device 32 is applied to the
input of an amplifier 34 that produces voltage at its output that is a
logarithmic function of the current, i, and is represented by the wave 36.
Because the body member absorbs less of the light of wavelength
.lambda..sub.1 than of light of wavelength .lambda..sub.2, the wave 36 has
a greater amplitude during the pulses 8' than during the pulses 10'. The
large peaks, 37, of the wave 36 occur when the blood pressure is at the
low point of its cycle because less blood is in the path 16 to absorb the
light. Conversely, the trough 39 of the wave 36 occurs at the high
pressure point in the heartbeat cycle when more blood is in the path 16 to
absorb the light. The time between the peaks 37 is therefore related to
the heart rate and is approximately one second. The frequency of the
pulses 4 is relatively much higher than illustrated.
In the interest of simplicity the wave 36 is drawn as though there were no
delay in the optical apparatus just described, but in practice, time must
be allowed for settling. For this reason the pulses 42' provided by the
sequence controller 2 on a lead 42 for the purpose of causing the sample
and hold circuit 38 to sample the voltage wave 36 are timed to occur
halfway through the pulses 8'. Similarly, the pulses 44' provided by the
sequence controller 2 on a lead 44 for the purpose of causing the sample
and hold circuit 40 to sample the voltage wave 36 are timed to occur
halfway through the pulses 10'. Each sample and hold circuit produces an
output voltage that is the same value as the value of one sample until the
next sample is taken. The output voltage of the sample and hold circuit 38
therefore, very closely corresponds to a wave drawn through the peaks of
the samples 42', and the output voltage of the sample and hold circuit 40
very closely corresponds to a wave drawn through the peaks of the samples
44'. The peak to peak amplitude of these waves, as indicated by the
brackets 42" and 44", is proportional to the amount of pulsatile blood.
The ratio of these amplitudes yields information as to the degree of
oxygen saturation.
It is an important part of this invention to provide means for passing the
A.C. components and blocking the D.C. components of the outputs of the
sample and hold circuits. For this purpose a bandpass amplifier 46 and a
capacitor 48 are coupled to the output of the sample and hold circuit 38
and a bandpass amplifier 50 and a capacitor 52 are coupled to the output
of the sample and hold circuit 40. The axis of these A.C. waves may be at
different D.C. levels depending on the effect of constant absorbers such
as hair, the venous blood flow and the pigmentation and thickness of the
skin through which the lights pass, but this has no effect on the A.C.
voltage waves at the output sides of the capacitors 48 and 52. The only
absorbers that have any effect are those contained in the arterial blood
flow, and they are the only ones of interest.
Any suitable means 54 and 56 can be respectively coupled to the capacitors
48 and 52 for the purpose of deriving an output signal, usually a voltage,
that varies as the average value of the A.C. waves above their lowest
values. One could also use a clamp circuit and an integrator, a peak to
peak detector, or even a full wave rectifier. As will be explained, it is
the ratio of these average values that yields the information as to the
desired degree of oxygen saturation SO.sub.2. The ratio can be determined
by simple division or by an analogue divider 58, and the degree of oxygen
saturation can be determined from a graph of the ratio vs. SO.sub.2, or,
if the ratio is digitized, it can be applied to a preprogrammed read only
memory or R.O.M. so as to yield the value of SO.sub.2 directly.
Other useful information can be derived by the apparatus just described,
e.g., the degree of perfusion can be attained directly from the output of
either averaging means 54 or 56 because the average values of either A.C.
wave are directly proportional to the arterial blood flow. In fact, if
only one of the channels A or B is used, the apparatus is the same as the
perfusion apparatus described in the U.S. Patent application Ser. No.
696,973, now Patent No. 4,109,643, entitled "Perfusion, Meter" filed on
June 17, 1976 in the name of Edwin B. Merrick. It is also possible to
determine the heartbeat rate by applying the A.C. output of either of the
bandpass amplifiers 46 or 50 to a trigger circuit 60 and coupling its
output to a cardiotachometer circuit 62.
Theory of Operation
In accordance with the Lambert-Beers Law, the ratio of light I of a single
wavelength that is transmitted by an optically absorbing material or
tissue to the incident light I.sub.o of the same wavelength is given by
the expression
I/I.sub.o =e.sup.-Ecd =e.sup.-A (1)
wherein c is the concentration of the absorber, d its thickness and E the
molecular extinction coefficient of the absorber for the particular
wavelength of light employed. It is convenient to refer to the product Ecd
as the absorption A.
By taking the log of both sides of (1) we obtain an expression for the
absorption A
-ln I/Io=Ecd=A. (2)
now, if the light is incident on a mixture of substances, each having its
own concentration and thickness, the total absorption is equal to the sum
of the absorptions of each substance, Thus, if CHb and CHbO.sub.2 are the
respective concentrations of the deoxyhemoglobin and the oxyhemoglobin in
the blood, Cn is the concentration of any fixed absorber such as skin or
hair, if dHb, dHbO.sub.2 and dN are the corresponding thicknesses of each;
and if EHb, EHbO.sub.2 and EN are the corresponding molecular extinction
coefficients, we can say for any particular wavelength of light that
-ln I/Io=EHb CHb dHb+EHbO.sub.2 CHbO.sub.2 dHbO.sub.2 +EN Cn dN=A. (3)
if blood is pulsing through the portion of a body member in the path 16 of
FIG. 1, the absorption due to the components of the blood will be changing
with time while in the absence of movement of the position of the body
member in the path 16 with respect to the optical system, the absorption
of all other components such as skin, hair, etc. remains constant and does
not change with time. Thus, by taking the derivative of (3) with respect
to time we obtain
-d/dt ln I/Io=d/dt EHb CHb dHb+d/dt EHbO.sub.2 CHbO.sub.2 DHbO.sub.2 +d/dt
EN CN dN=d/dt A, (4)
but the last term is zero because it is the derivative of a constant.
Furthermore
-d/dt ln I/Io=-d/dt ln I-d/dt ln Io (5)
and if Io is maintained constant, its derivative is also zero, so that
-d/dt ln I=d/dt EHb CHb dHb+d/dt EHbO.sub.2 CHbO.sub.2 dHbO.sub.2 =d/dt A.
(6)
equation (6) shows that, in the absence of field motion, the changes in the
logarithm of the transmitted light I are directly proportional to the
changes in the combined absorptions of the deoxyhemoglobin and the
oxyhemoglobin, or in other words, to the change in absorption of the
pulsatile or arterial blood. Thus, with a single wavelength of light we
can obtain an indication of the relative volume or perfusion of the blood.
Now let us assume that the photosensitive device 32 is a diode wherein the
current i generated therein by the transmitted light I can be expressed
for the selected wavelength of light as
i=kI (7)
where k is a constant related to the particular diode used and the
wavelength of light involved. When the current is passed through the
logging amplifier 34, we obtain an output voltage v that may be expressed
as
v=a ln i=a ln kI=a ln k+a ln I, (8)
where a is a constant relating the output voltage to the input current of
the amplifier 34. Now if we look at only the A.C. component of this
voltage as previously discussed in connection with the bandpass amplifier
46 and the capacitor 48, we have
.DELTA.v=a.DELTA. ln i=a.DELTA. ln k+a.DELTA. ln I (9)
and inasmuch as the change in the natural log of a constant k is zero we
can say that the A.C. component, .DELTA.v.sub.1 at the output of the log
amplifier 34 is
.DELTA.v.sub.1 =a.DELTA. ln i=a.DELTA. ln I=a.DELTA.A=a.DELTA. EHb CHb
dHb+a.DELTA. EHbO.sub.2 CHbO.sub.2 dHbO.sub.2. (10)
there are two important properties to note about this time-varying log
function; first, all fixed absorbers (skin, pigmentation, cartilage, bone,
hair, venous blood, etc.) have no influence on the final voltage; second,
all system gain factors drop out of the final equation if they can be
considered to multiply I.sub.o and are non-time varying.
Separation of Hb and HbO.sub.2
In order to determine the oxygen saturation, SO.sub.2, we need to separate
the absorption of Hb from that of HbO.sub.2. If we use two wavelengths of
light, e.g., red, I.sub.R at 660 nM and infrared I.sub.IR at 800 nM, for
example, then the ratio of the amplitude of the A.C. voltage, v, at the
output of channel A of FIG. 1 to the amplitude of the A.C. voltage at the
output of channel B may be expressed as the ratio of equations (10) for
each wavelength, or
##EQU1##
The concentration-distance product C.sub.b d.sub.b for the pulsatile blood
as a whole is equal to the sum of each of the absorption components. Thus,
if the only components are assumed to be H.sub.b and HbO.sub.2, their
percentages can be expressed as (100-x) and x. Thus:
C.sub.b d.sub.b =(C.sub.Hb d.sub.Hb +C.sub.HbO.sbsb.2
d.sub.HbO.sbsb.2)=(100-x)C.sub.b d.sub.b +xC.sub.b d.sub.b (12)
By using the above and observing that the a and C.sub.b d.sub.b cancel,
equation (11) becomes:
##EQU2##
Reference to FIG. 2 shows the molecular extinction coefficient for the
chosen wavelength of IR=800 nM to be 200 for both Hb and HbO.sub.2 but at
the chosen wavelength of R=660 nM, E.sub.RHb =900 and E.sub.RHbO.sbsb.2
=100. Substituting these values in (13) yields
##EQU3##
This is the equation of a straight line that is represented by the line 64
in the graph of FIG. 3. One end point of the line 64 is established by
making the concentration C.sub.HbO.sbsb.2 =to zero, i.e., let x=0. When
this is done (14) becomes
##STR1##
Then in order to establish the other end point let x=100 and
##EQU4##
Thus the change in voltage V.sub.R in channel A of FIG. 1 divided by the
change in voltage V.sub.IR in channel B of FIG. 1 yields a ratio shown on
the ordinate of the graph of FIG. 3, and the corresponding degree of
oxygen saturation SO.sub.2 is at the abcissa.
Use of 900 nM for IR
There is no known commercially available light emitting diode for producing
an IR of precisely 800 nM where the molecular extinction coefficients are
conveniently the same for both Hb and HbO.sub.2. Whereas light of this
wavelength could be produced by use of an incandescent light source and
filter at 800 nM, or by other means known to those skilled in the art, it
is simpler and far less expensive to use a commercially available source
of light having a wavelength of 900 nM. At this wavelength the molecular
extinction coefficient obtained from the graph for Hb in FIG. 2 is 200 and
the coefficient for HbO.sub.2 is 300. Substituting these values in (13)
yields
##EQU5##
Substitution of various values of x produces the dotted line 66 of FIG. 3.
Although this line is non-linear it is perfectly usable and has the same
properties as mentioned above.
Three Wavelength Oximeter
In describing two wavelength oximeter of FIG. 1, it was assumed that the
principle absorbers in the arterial blood flow were Hb and HbO.sub.2, but
it is sometimes of importance to know the percentage of a third absorber
such as carboxyhemoglobin, HbC, in which case three wavelengths of light
and three channels are used.
In FIG. 4, a sequence controller 68 provides pulses in sequence via the
leads 70, 72 and 74 to the gate electrodes of three FETs 76, 78 and 80
that have their source and drain electrodes respectively connected in
series with light emitting diodes 82, 84 and 86 between ground and a
source 88 of constant current so that the diodes respectively emit light
of three different wavelengths in sequence. These wavelengths may be 660
nM, 800 nM and 900 nM. Any suitable means such as a diffuser 89 can be
provided for directing light from each diode along a common path 90 to a
collection lens 92 for a photosensitive device 94.
The output current i generated by the photosensitive device 94 sequentially
corresponds in amplitude to the intensity of light that passes through the
body member from the diodes 82, 84 and 86. This current is applied to
means such as a logging amplifier 100 for producing an output voltage that
is a logarithmic function of the current. In particular, the grounded side
of the output of the photosensitive device is connected to the inverting
input of an operational amplifier 102 and the output current i is applied
to the non-inverting input. The base and collector electrodes of an NPN
transistor 104 are connected together and to the non-inverting input of
the amplifier, and the emitter of the transistor 104 is connected via a
resistor 106 to ground and by a resistor 108 to the output of the
operational amplifier 100. A diode could be substituted for the transistor
104, but the connection of the transistor in the manner described greatly
increases the accuracy of operation.
The output voltage of the logging amplifier 100 is applied to the source
electrodes of FETs 110, 112 and 114, that have their drain electrodes
respectively connected to capacitors 116, 118 and 120 so as to form sample
and hold circuits for the three channels A, B and C. The sequence
controller 68 provides short keying pulses via leads 122, 124 and 126 to
the gate electrodes of the FETs 110, 112 and 114 that respectively occur
near the middle of the pulses of light emitted by the diodes 82, 84 and
86. The timing of the keying pulses in this manner allows for the inherent
delay in the optical system and permits the output voltage of the logging
amplifier 100 to reach its maximum value in response to each pulse of
light before a sample is taken. When the FETs 110, 112, and 114 are placed
in a conducting condition by a keying pulse, the capacitors 116, 118 and
120 are respectively charged or discharged to the value of the output
voltage of the logging amplifier 100 that exists at that moment. After the
short keying pulse terminates, a capacitor retains its charge until the
next keying pulse is applied to the FET.
During the time between keying pulses the voltage across the capacitors
116, 118 and 120 is applied to the non-inverting inputs of operational
amplifiers 126, 128 and 130. The outputs of each amplifier is connected to
ground via series gain control resistors having their junction connected
to the respective inverting inputs of the amplifier. A.C. components of
the output voltages of the amplifiers 126, 128 and 130 are respectively
coupled via capacitors 132, 134 and 136 to the ungrounded ends of
resistors 138, 140 and 142. The ungrounded ends of these resistors are
respectively connected to the non-inverting inputs of operational
amplifiers 144, 146 and 148. The outputs of the operational amplifiers
144, 146 and 148 are respectively connected to ground via the parallel
combinations 150, 152 and 154 of a capacitor and resistor connected in
series with variable resistors 156, 158 and 160. With these connections
the operational amplifier 144, 146 and 148 function as bandpass amplifiers
with sufficient bandwidth to pass the frequency of the keying pulses
applied to the FETs and enough sidebands to define the pulses of arterial
blood such as are represented by the curve 36 of FIG. 1.
The outputs of the amplifiers 144, 146 and 148 are applied to an A to D
converter 162 which is controlled by timing pulses from the sequence
controller 68 on leads 164, 166 and 168 in such manner that the
digitalized outputs of the amplifiers 144, 146 and 148 appear on the leads
170, 172 and 174 respectively. These leads are connected to a
microprocessor 176 which is timed by pulses supplied to it by the sequence
controller 68 on a lead 178 so as to make calculations to be described
after each sequence of keying pulses have been applied to the gate
electrodes of the FETs 110, 112 and 114. Continuous readings of the
percent Hb, percent HbO.sub.2 and percent HbC appear at the output leads
178, 180 and 182 of the microprocessor 176, and readings of the heart rate
and perfusion index and oxygen saturation (SO.sub.2) appear on output
leads 184, 186 and 188 respectively. Additionally, cardiac output could be
calculated using a standard dye-dilution procedure.
Operation of FIG. 4
In this explanation the wavelengths of the light emitted by the light
emitting diodes 82, 84 and 86 are assumed to be as follows:
.lambda..sub.82 =660 nM
.lambda..sub.84 =800 nM
.lambda..sub.86 =900 nM
in discussing the operation of FIG. 1, it was shown that changes in the log
of the light I passing through the body member equals the sum of the
changes in absorption by the various absorbers. Therefore, when
considering three absorbers Hb, HbO.sub.2 and the carboxyhemoglobin HbC it
can be said
.DELTA. ln I.sub.82 =.DELTA.E.sub.Hb (.lambda..sub.82)C.sub.Hb d.sub.Hb
+.DELTA.E.sub.HbO.sbsb.2 (.lambda..sub.82)C.sub.HbO.sbsb.2
D.sub.HbO.sbsb.2 +.DELTA.E.sub.HbC (.lambda..sub.82)C.sub.HbC d.sub.HbC.
(18)
.DELTA. ln I.sub.84 =.DELTA.E.sub.Hb (.lambda..sub.84)C.sub.Hb I.sub.Hb
+.DELTA.E.sub.HbO.sbsb.2 (.lambda..sub.84)C.sub.HbO.sbsb.2
d.sub.HbO.sbsb.2 +.DELTA.E.sub.HbC (.lambda..sub.84)C.sub.HbC d.sub.HbC.
(19)
.DELTA. ln I.sub.86 =.DELTA.E.sub.Hb (.lambda..sub.86)C.sub.Hb d.sub.Hb
+.DELTA.E.sub.HbO.sbsb.2 (.lambda..sub.86)C.sub.HbO.sbsb.2
+.DELTA.E.sub.HbC (.lambda..sub.86)C.sub.HbC d.sub.HbC. (20)
the concentration-distance product C.sub.b d.sub.b for the pulsatile blood
as a whole is equal to the sum of the concentration-distance products for
each of the components. Thus, if the only components are assumed to be Hb,
HbO.sub.2 and HbC and if their percentages of the whole are respectively
X, Y and Z, the equations (18), (19) and (20) can be respectively written
as follows:
.DELTA. ln I.sub.82 =.DELTA.E.sub.Hb (.lambda..sub.82)XC.sub.b d.sub.b
+.DELTA.E.sub.HbO.sbsb.2 (.lambda..sub.82)YC.sub.b d.sub.b
+.DELTA.E.sub.HbC (.lambda..sub.82)Z.sub.C.sbsb.b.sub.d.sbsb.b. (21)
.DELTA. ln I.sub.84 =.DELTA.E.sub.Hb (.lambda..sub.84)XC.sub.b d.sub.b
+.DELTA.E.sub.HbO.sbsb.2 (.lambda..sub.84)YC.sub.b d.sub.b
+.DELTA.E.sub.HbC (.lambda..sub.84)Z.sub.C.sbsb.h.sub.d.sbsb.b. (22)
.DELTA. ln I.sub.86 =.DELTA.E.sub.Hb (.lambda..sub.86)XC.sub.b d.sub.b
+.DELTA.E.sub.HbO.sbsb.2 (.lambda..sub.86)YC.sub.b d.sub.b
+.DELTA.E.sub.HbC (.lambda..sub.86)Z.sub.Cbde. (23)
Substitution of the extinction coefficients of Hb, HbO.sub.2 and HbC for
the different wavelengths of light, as shown in the graphs of FIG. 2,
factoring out C.sub.b d.sub.b in the equations (21), (22) and (23) and
remembering that in the discussion of the operation of FIG. 1, it was
shown that the change in voltage, .DELTA.v, at the output of a channel,
like channels A, B or C in FIG. 4, was equal to .DELTA. ln I, it can be
said for similar reasons that the change in voltage .DELTA.v.sub.82 at the
output of channel A of FIG. 4 is equal to .DELTA. ln I.sub.82 ; the output
of channel B, .DELTA.v.sub.84 =.DELTA. ln I.sub.84, and the output of
channel C, .DELTA.v.sub.86 =.DELTA. ln I.sub.86 so as to yield
.DELTA. ln I.sub.82 =.DELTA.(900X+100Y+80Z)C.sub.b d.sub.b =.DELTA.v.sub.82
(24)
.DELTA. ln I.sub.84 =.DELTA.(200X+200Y+17Z)C.sub.b d.sub.b =.DELTA.v.sub.84
(25)
.DELTA. ln I.sub.86 =.DELTA.(200X+300Y+10Z)C.sub.b d.sub.b =.DELTA.v.sub.86
(26)
These simultaneous equations (24), (25) and (26) can be solved by use of
determinants as indicated by the equations (27), (28) and (29) below and,
as would be apparent to one skilled in the art, they can be evaluated by a
microprocessor 176.
##EQU6##
Effect of HbC on Two Wavelength Oximeter
As long as the only pulsatile light absorbers in the blood are Hb and
HbO.sub.2, the two wavelength oximeter of FIG. 1 yields accurate results,
but even when a third absorber such as HbCO is present in reasonable
quantities, this accuracy is still very good. Assume, for example, that
the two wavelengths of the light are used are 660 nM and 900 nM and that
the percentages of Hb, HbO.sub.2 and HbC are 44, 44 and 12 respectively.
Since by definition SO.sub.2 =HbO.sub.2 /(Hb+HbO.sub.2), SO.sub.2 is 50%.
In order to yield this result the ratio of the output voltages of the two
channels would have to be exactly 2, as seen from the graph 66 of FIG. 3.
This is what the ratio of the equations (24) and (26) would yield if the Z
terms were ignored, but since HbCO is present, the consequent inclusion of
the Z terms yields a ratio of 2.03. The coordinate value of SO.sub.2 for
this ratio from the curve 66 for the two wavelength oximeter is 51%, an
error of only 1% of maximum reading.
In the discussion of FIG. 1 the photosensitive device was assumed to have a
linear relationship between the light I falling on it and the current, i,
which it generated. However, devices having non-linear relationships can
be used. For example, let the device have a resistance that is affected by
the light in the following way.
R=AI.sup.k (30)
where R=the photoconductive resistance, I, the illumination and A and k are
constants depending on the material. A constant voltage E is applied
across the device so that the current i it produces is
i=(E/R)=(E/AI.sup.k) (31)
When the current, i, passes through a logging amplifier, it produces an
output voltage, v, that is proportional to the ln i, which is
v=ln i=ln E-ln A-K ln I (32)
both FIG. 1 and FIG. 4 include means for deriving a voltage v at the
outputs of the channels that is proportional to the pulsatile change in
the log of the current i with time so that
##EQU7##
or
.DELTA.v=.DELTA.ln i=-k.DELTA. ln I (34)
and inasmuch as it has previously been shown that .DELTA. ln
I=.DELTA.E(.lambda.)C.sub.b d.sub.b, the voltage at the output of each
channel corresponds to the absorption at the particular wavelength of
light to which it is responsive. Similarly, a phototransistor or any other
non-linear photoelectric device with a response of the form i=AI.sub.k
could be used.
It will be apparent to those skilled in the art that the sample and hold
circuits of FIGS. 1 and 4 are only one means for switching the electrical
signals produced by the photosensitive devices into separate channels, and
that the photosensitive devices themselves could perform this function. In
such event, a logging means would be necessary in each channel. In fact,
if desired, a separate logging means could be inserted in each channel in
the arrangement shown in FIGS. 1 and 4.
In FIG. 1 the logging amplifier 34, the sample and hold circuit 38, and the
bandpass amplifier 46 are means coupled to receive the electrical signals
provided by the photosensitive device 32 and for deriving therefrom an
output signal that is proportioned to the changes in the logarithm of the
detected light corresponding to one wavelength of light. Similarly, the
logging amplifier 34, the sample and hold circuit 40, and the bandpass
amplifier 50 are means coupled to receive the electrical signals provided
by the photosensitive device 32 and for deriving therefrom an output
signal that is proportional to the A.C. component of the logarithm of the
other transmitted wavelengths of light. When all of these components are
combined, or when separate logging means is included in each channel, as
described above, they form means coupled to receive the electrical signals
from the photosensitive means 32 and to derive separate output signals
that are respectively proportional to the alternating current component of
the logarithm of the electrical signals. The same understanding applies to
the arrangement of FIG. 4.
The measurement of cardiac output by indicator dilution techniques is based
on FICK's principle, namely, by injecting a known amount of indicator into
a stream of blood and measuring the concentration differences upstream and
downstream from the injection site, a measurement of a flow rate may be
calculated. The extension of this simple technique yields the familiar
cardiac output calculation. The importance of the measurement is not
questioned. However, the ideals of simplicity, speed, low cost, accuracy
and non-invasiveness have yet to be met.
As previously described in a multiple wavelength section, the presence of a
third absorber within the pulsatile blood stream may be dealt with by the
addition of a third wavelength. If this wavelength is properly chosen for
a particular dye, (cardio-green, methylene-blue, for example) the
indicator concentration may be determined by using the same incremental
absorbence techniques just described. Once this value is calculated the
familiar cardiac output calculation may be completed.
Some of the advantages of using incremental absorbence techniques versus
prior art for the dye concentration calculation are: that it eliminates
the need for a cuvette, the withdrawal of blood samples, the plumbing for
a flow-through cuvette and it is less invasive to the patient. Interfering
fixed absorbers such as bone, hair, skin tissue, pigmentation, etc. are
ignored as well as changes in the saturation of HbO.sub.2. Moment to
moment variations in the blood flow and blood velocity are automatically
accounted for.
* * * * *
|
|
 |
|
 |
< |
