United States Patent £191
`Po loge
`
`[54] ARTERIAL BLOOD MONITORING PROBE
`(75]
`Inventor:
`Jonas A. Pologe, Boulder, Colo.
`[73} Assignee: Ohmeda Inc., Murray Hill, N.J.
`[21] Appl. No.: 45,962
`Apr. 12, 1993
`[22} Filed:
`
`(SS]
`
`Related U.S. Application Data
`(63} Continuation-in-pan of Ser. No. 832,551, Feb. 7, t992.
`Int. Cl.' ................................................ A61B S/00
`[51]
`u.s. Cl ..................................... 128/633; 128/665;
`[52]
`356/41
`Field of Search ................................ 128/633-634,
`128/664-667; 356/39-41
`References Cited
`U.S. PATENT DOCUMENTS
`4,223,680 9/1980 Jobsis .................................. 128/633
`4,824,242 4/1989 Frick et al ...................... 128/666' X
`4,867,165 9/1989 Noller et at. .................... 128/666 X
`5;078,136 1/1992 Stone et at. ..................... 128/666 X
`5,127,406 7/1992 Yamaguchi ......................... 128/633
`
`[56]
`
`rllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll
`US005297548A
`5,297,548
`[Ill Patent Number:
`[45] Date of Patent: Mar. 29, 1994
`
`5,137,023 8/1992 Mendelson et at. ..•............. 128/633
`
`FOREIGN PATENT DOCUMENTS
`0303502 211989 Europw~ Pat. Off ..
`9004353 5/1990 PCT lnt'l Appl ..
`911l36 8/1991 PCT lnt'l Appl ..
`Primary Examiner-Angela D. Sykes
`Attorney, Agent, or Firm-Roger M. Rathbun; Larry R.
`Cassett; James M. Graziano
`[57]
`ABSTRACf
`This arterial blood monitoring system takes advantage
`of the basic statistical property that arterial blood con(cid:173)
`tains a plurality of dominant absorbers, whose measured
`light absorption spectra appear as a constant over a
`short interval of time. By measuring the transmitted
`light as it varies with arterial pulsation at selected wave(cid:173)
`lengths of light, over a common light path, the relative
`amount of these dominant absorbers in the arterial
`blood can noninvasively be detennined. To ensure the
`common light path, a sandwich constr!Uction light de(cid:173)
`tector is used.
`
`10 Claims, 4 Drawing Sheets
`
`Ill
`(~1120 ..£1030
`t~ EMITTER DRIVER
`'I.
`:.} 1"1 (103
`,..
`CIRCUIT
`c
`112b
`I
`~ETECTOR -
`105
`J
`
`101"'\
`104
`EMITTER
`
`1130
`113b
`
`113
`
`/102
`
`131
`
`CURRENT
`TO
`VOLTAGE
`CONVERTER
`
`SYNCHRONOUS -
`AMPLI· DEMODULA-
`FIER
`TION
`
`1327
`
`133,-/
`
`134,...
`
`I'" CHAN. I\
`CHAN.2
`
`CHAN. I
`
`DISPLAY MODULE
`
`PLETHYSMOGRAPHIC
`WAVEFORM
`
`114
`
`liSe
`
`Apple Inc.
`APL1020
`U.S. Patent No. 9,289,135
`
`001
`
`

`

`U.S. Patent
`
`Mar. 29, 1994
`
`Sheet 1 of 4
`
`5,297,548
`
`;eiOO
`
`102
`
`EMITTER DRIVER
`CIRCUIT
`
`131
`
`CURRENT
`TO
`VOLTAGE
`CONVERTER
`
`SYNCHRONOUS
`AMPLI- DEMODULA-
`FIER
`TION
`
`132
`
`137
`CHAN . I
`
`CHAN. 2 MUX
`
`ADC
`
`CHAN . n
`
`106
`
`CHAN. n
`
`CHAN . 2
`
`CHAN . I
`
`0::
`0
`~~
`icfW cu
`0 a:
`0..
`
`109
`
`DISPLAY
`DRIVER
`
`DISPLAY MODULE
`
`PLETHYSMOGRAPHIC
`WAVEFORM
`
`IMMI
`
`114
`
`115b
`
`115c
`
`FIG. I.
`
`002
`
`

`

`U.S. Patent ·
`
`Mar. 29, 1994
`
`Sheet 2 of 4
`
`5,297,548
`
`ARTERIAL PULSATION
`ABSORPTION
`ARTERIAL BLOOD
`ABSORPTION
`
`VENOUS BLOOD
`ABSORPTION
`
`OTHER TISSUE
`ABSORPTION
`
`TIME
`
`FIG. 2.
`
`z
`Q
`~
`0.. a:
`0
`(/)
`CD
`<t
`
`FIG.3.
`
`105
`
`-·
`
`113c
`
`003
`
`

`

`U.S. Patent
`
`Mar. 29, 1994
`
`Sheet 3 of 4
`
`5,297,548
`
`~1~~----~t------~
`
`----tRECEIVE DATA FROM PROBE AND
`STORE IN MEMORY
`
`410 "'\
`
`TRANSMIT DATA TO DISPLAY
`DRIVER TO PRODUCE
`WAVEFORM DISPLAY
`
`402~
`COMPUTE DIFFERENTIAL CHANGE IN
`ABSORPTION AT FIRST WAVELENGTH
`
`403---........_
`"-- COMPUTE DIFFERENTIAL CHANGE IN
`ABSORPTION AT SECOND WAVELENGTH
`
`~~~--------------~--------------,
`COMPUTE RATIO OF COMPUTED DIFFERENTIAL
`CHANGE IN ABSORPION FOR FIRST WAVELENGTH
`TO COMPUTED DIFFERENTIAL CHANGE IN
`ABSORPTION FOR SECOND WAVELENGTH
`
`405 ......---___
`
`~ t CQMPUTE tHb J
`
`406 ~ TRANSMIT COMPUTED tHb TO DISPLAY
`DRIVER TO PRODUCE NUMERIC OUTPUT
`
`. FIGURE
`
`4
`
`004
`
`

`

`U.S. Patent
`
`Mar. 29, 1994
`
`Sheet 4 of 4
`
`5,297,548
`
`501~
`RECEIVE DATA !'ROM PROBE UD
`! I
`8TORB IH JIEJIORY
`
`510,
`
`l
`
`TRANSMIT DATA TO DISPLAY
`DRIVER TO PRODUCE
`WAVEFORM DISPLAY
`
`502~
`COMPUTE DIFFERENTIAL CBAHGE IH
`ABSORPTION AT I'IRBT WAVELBHGTB
`
`503~
`COMPUTE DIFFERENTIAL CBAHGE ~
`ABSORPTIOH AT 8ECOHD ftVELDfG
`
`504
`
`.
`~ COMPUTE DIFl"BRBHTXAL CJIUGE IN
`r - - ABSORPTION AT THIRD RVBLDIQTB 1---
`
`505~
`
`50'~
`
`COMPUTE Sa02 FROM TBB
`HEAS'ORED DII'I'ERENTIAL
`CHANGE IN ABSORPTION
`AT ALL WAVELZHGTBS
`
`COMPUTE tHb PROM TD
`JIBAS'ORED Dil"l"BRDI'l'XAL
`CHANGE IN ABSORPTION
`AT ALL WAVELENGTHS
`
`507 \..
`
`! COMPUTE 02 ct !'ROM tHb UD 8&02
`
`_t
`
`t
`
`508 ---- TRANSMIT tHb, 8&02 AND 02 ct DATA
`
`TO DISPLAY DRIVER TO PRODUCE
`lf'OHBRIC OUTPUT
`
`FIGURE
`
`S
`
`005
`
`

`

`1
`
`5,297,548
`
`2
`tor. The greater the departure of the light beams from a
`common light path, the more significant the opportu·
`nity for the introduction of errors into the resultant
`measurements. This is especially true if multiple inde·
`5 pendent discrete light sources and multiple discrete
`light detectors are used in the probe, resulting in sepa·
`rate light transmission paths through the intervening
`appendage. The use of multiple light detectors, each
`sensitive to different wavelength regions, becomes a
`10 necessity if the wavelengths of light selected are far
`apart in wavelength, since there does not exist a single
`light detector device that can detect a wide bandwidth
`of light with significant speed, sensitivity and an accept·
`ably flat response. Therefore, existing probe designs can
`introduce errors into the measurements by their inabil·
`ity to transmit a plurality of light beams substantially
`along a common light path through the arteriolar bed of
`the appendage being monitored.
`
`ARTERIAL BLOOD MONITORING PROBE
`
`CROSS-REFERENCE TO RELATED
`APPLICATIONS
`This application is a continuation-in-part of U.S. pa·
`tent application Ser. No. 071832,SS1, titled "Improved
`Arterial Blood Monitoring System", filed Feb. 7, 1992,
`pending
`
`FIELD OF THE INVENTION
`This invention relates to non-invasive photoplethys·
`mographic measurement of blood analytes and, in par·
`ticular, to a probe for use in an arterial blood monitoring
`system to more accurately measure the change in inten· 15
`sity of the light transmitted through the arterial blood of
`a patient.
`
`PROBLEM
`SOLUTION
`It is a problem in the field medical monitoring equip· 20
`The above described problems are solved and a tech·
`ment to accurately measure various parameters of arte·
`rial blood in a noninvasive manner. For example, the
`nical advance achieved in the field by the probe for an
`oxygen saturation (So (h) of the hemoglobin in arterial
`blood is determined by the relative proportions of oxy·
`arterial blood monitoring system that creates a single
`genated hemoglobin and reduced hemoglobin in the 25 light path through an appendage to noninvasively mea·
`sure and calculate characteristics of arterial blood. This
`arterial blood. A pulse oximeter system noninvasively
`determines the oxygen saturation of the hemoglobin by
`arterial blood monitoring system probe talces advantage
`measuring the difference in the light absorption of these
`of the basic statistical property that arterial blood con·
`two forms of hemoglobin. Reduced hemoglobin absorbs
`tains a plurality of dominant absorbers, whose measured
`more light in the red band (600-800 nm) than does oxy· 30 light absorption spectra appear as a constant over a
`short interval of time. The arterial blood characteristics
`hemoglobin while oxyhemoglobin absorbs more light in
`the near infrared band (800-1000 nm) than does reduced
`to be measured are empirically related to the changes in
`hemoglobin.
`the measured light transmission through the plurality of
`The pulse oximeter includes a probe that is placed in
`dominant absorbers as a function of the changes in arte·
`contact with the skin, either on a flat surface in the case 35 rial blood volume at the probe site. By measuring the
`of reflectance probes or across some appendage in the
`transmitted light as it varies with a.rterial pulsation at a
`case of a transmission probe. The probe contains two
`plurality of selected wavelengths of light, over a single
`light emitting diodes, each of which emits a beam of
`common light path, the relative amount of these domi·
`light at a specific wavelength, one in the red band and
`nant absorbers in the arterial blood can noninvasively
`one in the infrared band. The magnitude of red and 40 be determined.
`By selecting one wavelength of light around 1270nm,
`infrared light transmitted through the intervening ap·
`pendage contains a non-pulsatile component which is
`where water has a measurable extinction and second
`influenced by the absorbency of tissue, venous blood,
`and third wavelengths at about 660 nm and 940 nm, a
`capillary blood, non-pulsatile arterial blood, and the
`direct relationship between the transmitted intensities at
`intensity of the light source. The pulsatile component of 45 these three wavelengths and the arterial hemoglobin
`concentration exists and can be calculated. The accu·
`the received signals is an indication of the expansion of
`the arteriolar bed in the appendage with arterial blood.
`rate detection of these three wavelengths of light is
`The effects of different tissue thicknesses and skin pig-
`accomplished by the use of two different light detec·
`mentation in the appendage can be removed from the
`received signals by normalizing the change in intensity 50 tors. To avoid the problem of different light paths
`of the received signal by the absolute intensity of the
`through the intervening appendage, a sandwich or lay·
`received signal. Taking the ratio of the mathematically
`ered detector design is used in the probe. The light
`processed and normalized red and infrared signals re·
`detector consists of a multiple layer element that con·
`suits in a number which is theoretically a function of
`tains a germanium photodiode placed under, and coinci·
`only the concentration of oxyhemoglobin and reduced 55 dent with, a silicon photodiode. For the wavelengths of
`hemoglobin in the arterial blood. This assumes that
`light shorter than approximately 1000 om, the silicon
`oxyhemoglobin and reduced hemoglobin are the only
`photodiode receives the incident light and produces a
`substantial absorbers in the arterial blood.
`signal indicative of the intensity of the received light.
`Above this wavelength, the. silicon pho~odi~e becomes
`The amplitude of the pulsatile component is a very
`small percentage of the total signal amplitude and de· 60 transparent and the germaruum photodaode packs up the
`incident light. Thus, the light from the three light
`pends on the blood volume change per pulse and the
`oxygen saturation (S00 2) of the arterial blood. The
`sources is transmitted through the tissue along substan·
`received red and infrared signals have an exponential
`tially identical light paths to be detected by the coinci·
`relationship to the path length of the arterial blood. The
`dent light detectors at exactly the same "exit area",
`photoplethysmographic measurement of these analytes 6S regardless of wavelength. By constraining the detected
`· light to traverse one path through the tissue, regardless
`is predicated on the assumption that the light beams
`of wavelength, this apparatus avoids the inaccuracies
`from the two light sources follow identical paths
`through the intervening appendage to the light detec-
`caused by sampling different cross-sections of tissue, as
`
`006
`
`

`

`3
`with two or three discrete light detectors mounted side
`by side.
`
`5,297,548
`
`BRIEF DESCRIPTION OF THE ORA WING
`FIG. 1 illustrates in block diagram form the overall
`architecture of the arterial blood monitoring system and
`the probe of the present invention;
`FIG. 2 illustrates in graphical form the various com(cid:173)
`ponents of the input signal from the probe;
`FIG. 3 illustrates a cross-section view of the light
`detector used in the probe of the present invention;
`FIG. 4 illustrates in flow diagram form the opera(cid:173)
`tional steps taken by a two wavelength arterial blood
`monitoring system to measure selected components in
`15
`arterial blood; and
`FIG. S illustrates in flow diagram form the opera- .
`tional steps taken by a three wavelength arterial blood
`monitoring system to measure selected components in
`arterial blood.
`
`DETAILED DESCRIPTION
`An arterial blood monitoring system takes advantage
`of the basic statistical property that arterial blood con(cid:173)
`tains a plurality of dominant absorbers, whose measured
`light absorption spectra appear as a constant over a 25
`short interval of time. The arterial blood characteristics
`to be measured are empirically related to the changes in
`the measured light transmission through the plurality of
`dominant absorbers as a function of the changes in the
`arterial blood volume at the probe site. Therefore, by 30
`measuring the transmitted light as it varies with arterial
`pulsation, at selected wavelengths, the relative amount
`of these dominant absorbers in the arterial blood can
`noninvasively be determined. A single probe can be
`used to generate the plurality of wavelengths of light, _35
`therefore simplifying the arterial blood monitoring sys(cid:173)
`tem.
`
`4
`The probe 101 consists of an exterior housing 104 that
`applies the active elements of the probe 101 to the tissue
`under test, such as a fmger 105, containing an arterial
`blood flow that is to be monitored. Included within
`S housing 104 is a plurality (at least two) of light emitting
`devices 111, 112 and at least one corresponding light
`detector 113.
`Emitter driver circuit 131 produces the analog drive
`signals to activate light emitting devices 111, 112 in
`10 probe 101. These analog drive signals are carried over
`cable 103 to probe 101. To measure the concentration of
`total hemoglobin (tHb), oxygen saturation (Sg(h), or
`other blood analytes, in arterial blood, the concentra-
`tion of several dominant absorbers contained in the
`arterial blood must be measured. In particular, for the
`measurement of total hemoglobin (tHb), concentration
`of the water and hemoglobin components of the arterial
`blood must be measured. The light emitting devices 111,
`112 each produce an output light beam of predeter-
`20 mined wavelength which is directed at the fmger 105
`enclosed by housing 104. In this embodiment, light
`emitting device 111 is selected to produce a beam of
`light at approximately 810 nm, which wavelength is
`substantially isobestic to the oxygenated and deoxygen(cid:173)
`ated components of the hemoglobin in the arterial blood
`(that is, the extinction coefficients of the oxygenated
`and deoxygenated hemoglobin are substantially identi(cid:173)
`cal). Light emitting device 112 is selected to produce a
`beam of light at approximately 1270 nm. The selection
`of these two wavelengths is such that water is transpar(cid:173)
`ent at the first wavelength oflight (810 nm) but detected
`at the second (longer) wavelength of light (1270 nm). ln
`addition, these wavelengths are such that the extinction
`coefficients of the two components (water and hemo(cid:173)
`globin) differ at the first wavelength of light. Further, at
`both wavelengths the two species of hemoglobin are
`substantially isobestic in extinction but not transparent.
`The light detector 113 monitors the level oflight that
`is transmitted through or reflected from finger 105. The
`Definition of Terms
`lo=The intensity of the beam of light at a given wave- 40 analog data signals produced by light detector 113 in
`length incident on the tissue-under-test, where the
`response to the received beams of light are received
`wavelength is denoted by the subscript.
`from probe 101 over conductors 103 and flltered by
`I= The instantaneous value of the intensity of the light
`. analog hardware 132- 134 in probe interface circuit 102.
`received by the detector. The light is at a given wave-
`The input analog data from probe 101 may be decom(cid:173)
`length, which wavelength is indicated by a subscript. 45 posed into its non-pulsatile and pulsatile sub-elements in
`£=The extinction coefficient of light by a given sub-
`probe interface circuit 102 in order to provide accurate,
`stance (indicated by a superscript) at a given wave-
`high resolution, measurements of these components.
`The pulsatile component typically represents anywhere
`length (indicated by a subscript).
`C=The concentration of a given substance (indicated
`from 0.05% to 20% of the total input signal and the
`50 decomposition of the input signal into pulsatile and
`by a superscript).
`L = The pathlength of a given substance (indicated by a
`non-pulsatile components permits accurate analog to
`superscript).
`digital conversion of even the smallest of these pulsatile
`tHb = Total hemoglobin measured in arterial blood.
`components.
`Usually expressed in terms of grams per deciliter.
`In order to distinguish between the light beams pro-
`0 = Used as a superscript to represent oxyhemoglobin. 55 duced by ftrst 111 and second 112 light emitting de-
`R = Used as a superscript to represent reduced hemo-
`vices, these light emitting devices 111, 112 are modu-
`globin.
`lated in a manner to allow the output of the light detec-
`W = Used as a superscript to represent water.
`tor 113 to be synchronously demodulated. Ambient
`t = Used as a superscript to represent the combination of
`light, being unmodulated, is easily eliminated by the
`oxyhemoglobin and reduced hemoglobin.
`60 demodulator process.
`
`System Architecture
`FIG. 1 illustrates in block diagram form the overall
`architecture of the arterial blood monitoring system 100
`and the probe 101 of the present invention. The arterial 65
`blood monitoring system 100 consists of a probe 101
`connected to probe interface circuit 102 by means of a
`set of electrical conductors 103 and connector 103a.
`
`Signal Components
`FIG. 2 illustrates in graphical form (not to scale) the
`various components of the total absorption produced by
`finger 105. The light detector output signal, high where
`absorption is low and visa versa, consists of a large
`magnitude non-pulsatile component and a small magni(cid:173)
`tude pulsatile component. The non-pulsatile component
`
`007
`
`

`

`5,297,548
`
`6
`scaling amplifiers 135 such that they can be converted,
`with optimal resolution, to a digital equivalent. All
`channels output by scaling amplifiers 13S are then si(cid:173)
`multaneously· sampled by the sample/hold circuitry
`5 136a, 136b, . . . 136n. The sampled data is passed a chan(cid:173)
`nel at a time via multiplexer 137 to the analog to digital
`converter 138. From there the data, now in digital form,
`is sent on to data processing circuit 107 where it is
`stored in memory 106 for processing. The digital data
`10 represents the substantially simultaneously sampled
`amplitudes of the received light intensities from each of
`the wavelengths used at a sampling frequency of typi(cid:173)
`cally 30Hz or greater. These data values are referred to
`as l~t h ... . IN, where the subscript indicates the given
`wavelength. I, then indicates the received light inten(cid:173)
`sity at any given wavelength.
`
`5
`represents light remaining after absorption due to a
`combination of venous blood, cutaneous tissue, bone,
`and constant arterial blood while the small pulsatile
`component is caused by the light absorption due to
`pulsatile arterial blood flow that is to be measured.
`Following synchronous demodulation, the data signals
`produced by light detector 113 and transmitted to probe
`interface circuit 102 consist of a series of data points that
`are digitized and stored in memory 106. Since the fJTSt
`111 and second 112 light emitting devices are sampled
`simultaneously and in. rapid succession, these digitized
`data points consist of a plurality of sets of measure(cid:173)
`ments, with one set corresponding to samples of the
`light beam intensity at a first wavelength, the other set
`corresponding to samples of the light beam intensity at 15
`a second wavelength, and, in some schemes, a third set ·
`corresponding to the intensity of the ambient light.
`Ideally, in pulse oximeter systems red and infrared
`wavelengths of light are used and the ratio of the nor(cid:173)
`malized derivative (or logarithm) of the red intensity to 20
`the normalized derivative (or logarithm) of the infrared
`intensity is a constant. This constant is indicative of the
`partial oxygenation (Sa02) of the hemoglobin in the
`arterial blood flow. It is obvious that this ratio changes
`as Sa02 changes but, for a short interval with rapid 25
`enough sampling rate, the ratio remains constant.
`
`Data Processing Circuit
`In a two wavelength system, data processing circuit
`107 computes a ratio from the digital amplitude data
`measured at each wavelength oflight. In particular, this
`process used by data processing circuit 107 is illustrated
`in flow diagram form in FIG. 4. At step 401, data pro(cid:173)
`cessing circuit 107 receives a set of digital input data
`indicative of the measured intensity oif light at both
`wavelengths, as received by light detector 113. Data
`processing circuit 107 at step 410 transmits the received
`set of data to display driver 109 for display in graphical
`form on display 114. The displayed waveform repre(cid:173)
`sents the pulsatile component of the arterial blood. Data
`processing circuit 107 also stores the received set of
`data in memory 106 and uses this set of data and the last
`most recently received set of data to compute at steps
`402 and 403 the differential change in absorption of the
`arterial blood in finger lOS at the fJTSt and second se(cid:173)
`lected wavelengths of light, respectively. The differen(cid:173)
`tial change in absorption at wavelength n is computed
`by data processing circuit 107 as:
`
`Probe Interface Circuit
`The actual analog data received by the probe inter(cid:173)
`face circuit 102 can include a fairly significant noise 30
`component which is caused by a number of sources
`including motion of finger lOS, the introduction of am(cid:173)
`bient light into housing 104, and various sources of
`electrical noise. These noise components skew the val(cid:173)
`ues of either or both of the magnitudes measured in each 3.S
`set of data points destroying the correct relationship
`between the red and infrared signals. Existing pulse
`oximeter circuits make use of various filtering tech(cid:173)
`niques to minimize the impact of noise on the Sa02
`value measured by the system. This flltering circuitry 40
`and software/algorithms are analogous to that used in
`the arterial blood monitoring system 100 and are there-
`. .
`·
`fore not described in detail herein.
`.
`.
`.
`1
`Probe interface circuit 102 includes emitter driver
`Because di, IS a mathematlca construct, 1t JS approxJ·
`circuit 131 that is capable of driving light emitting de- 45 mated in arterial blood monitoring syste m 100 by .AI,,
`where .AI, is the difference between two consecutively
`vices 111, 112 such that the light beams produced tra-
`received I, values. Only AI values that are caused by a
`verse finger lOS and sufficient light intensity is incident
`on light detector 113 to produce data indicative of the
`small but non zero change in path length through finger
`lOS are used and therefore .AI, can also be a longer
`light absorption of the dominant absorbers in arterial
`blood. The data produced by light detector 113 (voltage 50 interval oftime if necessary to obtain a sufficient change
`equivalent of the received light intensities) at each
`in received intensity of the beam of light. The I, value
`used in equation 1 is the average of the two successively
`wavelength is kept distinct and can be processed inde-
`pendently. This can be done by any of the many
`received I, values used to compute .AI,.
`schemes presently in use for pulse oximetry, such as
`In a two wavelength system, a final ratio is then cal-
`time division multiplexing, or frequency division multi- 55 culated by data processing circuit 107 at step 404 as:
`plexing.
`The light received from finger lOS is converted to an
`equivalent current signal by the photodiodes of light
`detector 113, and then converted to a voltage signal by
`the current to voltage converter 132. The data is then 60
`amplified by amplifier 133, and demultiplexed via syn(cid:173)
`chronous demodulation circuit 134. The demultiplexed
`data comprises analog voltage signals applied to leads
`CHAN 1, CHAN 2 ... CHAN n representative of the
`intensity of the received light at each of the wave- 65
`lengths of light produced by light emitting devices 111,
`112, respectively. The voltage signals on leads CHAN
`1, CHAN 2 are then scaled (further amplification) by
`
`where the data values used to compute dAt are from the
`same points in time as the data values used to compute
`dA2.
`This ratio is then used in a calibration equation by
`data processing circuit 107 at step 405 to relate the R
`value to a specific blood analyte value. For example,
`when measuring total hemoglobin, the calibration equa(cid:173)
`tion is approximated by a second order polynomial of
`the form:
`
`(1)
`
`(2)
`
`008
`
`

`

`7
`
`5,297,548
`
`(3)
`
`8
`blood monitoring system 100 to read the differential
`change in absorption (dA values) as accurately as possi(cid:173)
`ble. This leaves only the values of the differential path
`lengths dL as unknowns. With two equations, the two
`5 dL values can be uniquely and simply solved for.
`Writing the proportion of hemoglobin in the arterial
`blood as:
`
`Where A, B, and C are constants that depend on the
`specific wavelengths of light used.
`The tHb value is then output by data processing cir(cid:173)
`cuit 107 at step 406 to display driver 109 (and/or hard·
`copy) to display in human-readable form on display 115
`a numeric value of the concentration of total hemoglo·
`bin in the arterial blood of finger 105. Processing then 10
`returns to step 401.
`
`Theory
`This device is based on the theory that:
`
`(4) 15
`
`A differential change in absorption at a given wave(cid:173)
`length n to a given substance (dA,,S), is equal to the
`extinction of that substance (Ens) times the concentration
`(CS) of that substance times the differential change in 20
`pathlength of that substance (dU).
`Further the differential change in absorption can be
`defmed as:
`
`din
`dAn=/;"'
`
`(S) 25
`
`Note that no measurement of the incident light inten·
`sity, Io, is required to measure the differential change in 30
`absorption dA. However, samples of I, must be taken
`sufficiently close in time so that t..I, represents a good
`mathematical approximation of di,.
`To determine the relative proportions of two domi(cid:173)
`nant absorbers, in this case water and hemoglobin, one 35
`chooses two wavelengths of light at which the two
`absorbers have extinctions, such that the following set
`of simultaneous equations has a unique solution for all
`possible concentrations and pathlengths of the two ab·
`sorbers.
`
`40
`
`(6)
`
`(7}
`
`In this system of equations it is assumed that the only 45
`components which change in pathlength are those of
`the arterial blood. Further it is assumed that the primary
`absorbers are those of water and hemoglobin where the
`hemoglobin species in the blood are essentially only
`those of oxyhemoglobin and reduced hemoglobin. 50
`Choosing a wavelength of light that represents an
`isobestic point for the two species of hemoglobin, such
`as 804 manometers, minimizes the effects of changes in
`oxygen saturation on the total hemoglobin readings.
`Notice that in equations 6 and 7 above, the concentra- 55
`tion term expressed in equation 4 has been eliminated.
`By viewing the optical system as compartmentalized,
`that is, looking at the tissue under test as one in which
`the light first passes through 100% skin tissue, followed
`by 100% venous blood, followed by 100% arterial he- 60
`moglobin, followed by 100% water, and so on, the
`concentration terms expressed by equation 4 are actu·
`ally constants. Thus, beginning with equation 6, the
`extinction coefficients are meant to represent the combi·
`nation of the actual extinction coefficient and the actual 65
`concentration for 100% of any given absorber.
`In the system of equations (Equations 6, 7) the extinc·
`tion E values are constants and it is the job of the arterial
`
`Proponion Hb = _ _:d:;L,_'(cid:173)
`dL1 + dL"'
`
`(8)
`
`While this proportion is not directly the tHb, it is di·
`rectly related to it. And while this relationship could be
`theoretically derived, an empirical relationship (as de·
`fined in equation 3) is measured instead. This is neces·
`sary due to several ways in which the true optical sys(cid:173)
`tem of living tissues and realistic optical elements devi·
`ate from the exact theoretical model developed here.
`Equation 3 is therefore referred to as the calibration
`equation and its coefficients A, B, and C, are experimen·
`tally derived via clinical testing. The coefficients are
`then installed in the arterial blood monitoring system
`software. It should be noted that these coefficients dif·
`fer for different wavelength emitters.
`The wavelengths of light produced by light emitting
`devices 111, 112 are also selected so as to optimize the
`performance of the entire electro optical system: low
`enough light absorption so that sufficient optical signal
`is received by light detector 113 and high enough light
`absorption so that there is an appreciable change in light
`absorption over the range of physiological changes in
`pathlength caused by the pulsation of arterial blood.
`Typical wavelengths of light selected for a realization
`of this system are 810 nm and 1270 nm, however many
`wavelength combinations meeting the above criteria
`can be used.
`
`Probe
`Probe 101 contains a minimum of two lights emitting ·
`devices 111, 112, each of which produces a beam of
`light centered about a selected wavelength (810 nm and
`1270 nm, respectively). Probe 101 also contains light
`detector 113 capable of receiving the emitted wave·
`lengths of light. In the present implementation, the light
`detector 113 consists of a multiple layer element, shown
`in additional detail in FIG. 3, that contains a germanium
`photodiode 113b placed under a silicon photodiode
`113a. For the wavelengths of light shorter than approxi·
`mately 1000 nm, the silicon photodiode 113a receives
`the incident light. Above this wavelength, the silicon
`photodiode 113a becomes transparent and the germa·
`nium photodiode 113b picks up the incident light. Probe
`101 includes a cable 103 and connector 103a for trans·
`mitting and receiving signals between probe 101 and
`probe interface circuit 102. Probe 101 is positioned on
`the tissue either in the transmission mode: light emitting
`devices 111, 112 on one side and light detector 113 on
`the other side of fmger 105, earlobe, toe or other appro(cid:173)
`priate site through which light can be received by the
`light detector 113 at acceptable signal levels; or in the
`reflectance mode: in which the light emitting device
`111, 112 and light detector 113 are placed on the same
`side of the tissue under test, such as the forehead or
`forearm.
`
`009
`
`

`

`5,297,548
`
`10
`(02Hb), reduced hemoglobin (RHb), and water (H20).
`
`(10)
`
`mined, and curve fit, experimentally. The empirical
`curve fit also compensates for the differences between
`the theoretical models and the actual optical systems.
`
`(11)
`
`For a three wavelength system, with the subscripts 1, 2,
`and 3 indicating the specific wavelengt!hs used we can
`write following system of equations
`
`9
`Combination tHb Monitor and Pulse Oximeter
`The methodologies for pulse oximetry are well
`known. The method of obtaining tHb noninvasively
`and in real time has been disclosed above. The arterial s
`blood monitoring system of the present invention can
`combine the two technologies to create a device for
`measurement of both parameters. tHb is an interfering
`Note that this equation shows only that !the total heme-
`substance in the measurement of Sa02 by the present
`technologt·es. By "interfering substance" it is meant that 0 globin is proportional to this ratio of path length
`1
`changes, not equal to it. This is due, at Jeast in pan, to
`variations in tHb cause variations in the SaOl as ready
`by a pulse oximeter. These variations in Sa02 are corre-
`the fact the tHb is measured in terms of grams/deciliter
`lated to, but not corrected for, the tHb level. A device
`of whole blood and this equation is a ratio of path
`capable of measuring tHb can therefor provide a means
`lengths. There is a one to one correspondence between
`for eliminating the error it causes in determining Sa02. IS this ratio of path lengths and the tHb which is deter(cid:173)
`The s