WO2015175360A1

WO2015175360A1 - Pulseless oximeter to estimate arterial oxygen saturation noninvasively in patients with weak or absent pulses

Info

Publication number: WO2015175360A1
Application number: PCT/US2015/030057
Authority: WO
Inventors: Thomas K. Aldrich
Original assignee: Montefiore Medical Center
Priority date: 2014-05-13
Filing date: 2015-05-11
Publication date: 2015-11-19

Abstract

Methods and systems are provided for estimating arterial oxygen saturation (Sa02) in patients in whom pulsatile blood flow is reduced or absent.

Description

PULSELESS OXIMETER TO ESTIMATE ARTERIAL OXYGEN SATURATION

NONINVASIVELY IN PATIENTS WITH WEAK OR ABSENT PULSES

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 61/992,292, filed on May 13, 2014, the contents of which are herein incorporated by reference in their entirety.

BACKGROUND OF THE INVENTION

[0002] Throughout this application various publications are referred to in parentheses. Full citations for these references may be found at the end of the specification. The disclosures of these publications are hereby incorporated by reference in their entirety into the subject application to more fully describe the art to which the subject invention pertains.

[0003] Arterial oxygen saturation (Sa0₂) refers to the concentration of oxygen in arterial blood. Arterial oxygen saturation (Sa0₂) estimated by pulse oximetry is referred to as Sp02. Pulse oximetry estimates arterial oxygen saturation by taking a ratio of pulsatile changes in attenuance (absorbance plus scatter) of a fingertip at two wavelengths of light, typically 660nm (red) and 905nm (infrared), selected for their differing absorption characteristics by oxy- and reduced hemoglobins (Weiben, 1997; Zjilstra, 1991). By measuring pulsatile changes in attenuance, contributions from all components of the finger, with the exception of the pulsing arterial blood, are factored out. For each wavelength, changing attenuance is determined by (1) pulsatile changes in the volume of blood in the light path, (2) hemoglobin concentration, (3) relative concentrations of oxy- and reduced hemoglobins, and (4) the attenuation coefficients of oxy- and reduced hemoglobins at the relevant wavelength. By taking a ratio of results at two wavelengths, the pulse volume, and hemoglobin concentration terms cancel out, allowing saturation to be estimated by comparison of the ratio with empirically-derived standard data (Weiben, 1997).

[0004] Pulse oximetry fails when the pulse volume is too small to achieve acceptable signal to noise ratios. Common clinical scenarios of pulse oximetry failure include patients with peripheral vascular disease, patients in circulatory shock, and patients on continuous flow left ventricular assist devices (LVADs). For such patients, if arterial oxygenation must be assessed, the only option is arterial puncture for arterial blood gas measurement, which is painful, carries a small risk of arterial injury, and provides only a "snapshot" assessment of oxygenation, not the more useful continuous monitoring capability of pulse oximetry. Therefore, a noninvasive oximeter that does not require pulsatile flow will have a valuable role in assessment of such patients.

[0005] The present invention addresses the need for improved methods and apparatus for measuring arterial saturation noninvasively in patients in whom pulsatile blood flow is reduced or absent.

SUMMARY OF THE INVENTION

[0006] The invention provides apparatus and techniques of estimating arterial blood oxygenation (percentage of oxyhemoglobin or saturation of arterial blood hemoglobin with oxyhemoglobin) in persons, in particular in persons with insufficient pulse volume for standard pulse oximetry.

[0007] The present invention provides methods of estimating arterial oxygen saturation (Sa0₂) in a patient in whom pulsatile blood flow is reduced or absent, comprising: alternately transilluminating a tissue of the patient with two or more wavelengths of light; momentarily occluding arterial blood flow to the tissue of the patient, and then releasing occlusion of arterial blood flow to the tissue; measuring change in transmission separately for each of the two or more wavelengths of light across the tissue during occlusion of arterial blood flow to the tissue and after release of occlusion of arterial blood flow to the tissue, where light levels at each wavelength at points of time after release of occlusion are expressed as a fraction of the maximum light level that occurs during occlusion, yielding a relative transmittance (T) level for each wavelength of light; calculating arterial blood attenuance (A) for each wavelength at each time point, where A = log (1/T); calculating an average arterial blood attenuance ratio for the two or more wavelengths; and estimating arterial oxygen saturation (Sa0₂) in the patient by comparing the average arterial blood attenuance ratio with data from subjects with known Sa0₂ or with data from calibrating devices.

[0008] The invention also provides systems for estimating arterial oxygen saturation (Sa0₂) in a patient in whom pulsatile blood flow is reduced or absent, comprising: one or more devices for emitting two or more wavelengths of light; one or more light detecting devices; and one or more computing devices comprising one or more processors, a memory unit, a display device, and a computer-readable storage medium including computer- readable code that is read by the one or more processors to perform a method comprising the steps of: alternately transilluminating a tissue of the patient with two or more wavelengths of light; measuring change in transmission separately for each of the two or more wavelengths of light across the tissue during occlusion of arterial blood flow to the tissue and after release of occlusion of arterial blood flow to the tissue, where light levels at each wavelength at points of time after release of occlusion are expressed as a fraction of the maximum light level that occurs during occlusion, yielding a relative transmittance (T) level for each wavelength of light; calculating arterial blood attenuance (A) for each wavelength at each time point, where A = log (1/T); calculating an average arterial blood attenuance ratio for the two or more wavelengths; and estimating arterial oxygen saturation (Sa0₂) in the patient by comparing the average arterial blood attenuance ratio with data from subjects with known SaC>₂ or with data from calibrating devices.

BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1. Example of manual occlusion of radial and ulnar arteries.

[0010] Figure 2. Raw photoplethysmograms. Red and infrared photoplethysmograms collected from one normal subject, during quiet breathing on room air, followed by a 5 second occlusion of radial and ulnar arteries, during which light transmission increases, because venous blood drains, while there is no arterial inflow. At the 17 second time point, the occlusion was released, causing decrease in transmitted light due to arterial blood inflow. The data identified by the oval are selected for further analysis.

[0011] Figure 3. Transmittances and attenuances. For the one second after release of occlusion, raw photoplethysmographic data are expressed as fractions of the maximal level (transmittance (T), declining due to influx of arterial blood, top two curves). Each data point is then converted to attenuance (A = log(l/T) of the inflowing arterial blood, bottom two curves).

[0012] Figure 4. Calculating Red/Infrared changing attenuance ratio (R/IR). Red attenuance is plotted against infrared attenuance for the data from nadir to peak attenuance. The slope of the resulting line represents average R/IR ratio, in this case 0.566.

[0013] Figure 5. Sp02 or Sa02 plotted against red/infrared (R/IR) changing attenuance ratio. Closed symbols delineate median R/IR and Sp0₂ data from five normal subjects breathing various concentrations of oxygen. The line and the equation in the box represent the regression equation for the normal subjects. Sp0₂ was highly significantly correlated with R/IR for the normal subjects. The open symbols represent nine data points (arterial oxygen saturation (Sa0₂) measurements from arterial blood and median R/IR values) from seven LVAD patients. Among the LVAD patients, «Sa0₂ was strongly correlated with R/IR (R²=0.75, p<.003 for both intercept and slope). The intercept was identical to that observed in normals (1 11). Slope was slightly shallower (24.8). The average error (the difference between SaC>₂ measured in blood and that predicted by the regression equation) was 1.1 percentage points.

DETAILED DESCRIPTION OF THE INVENTION

[0014] The present invention provides a method of estimating arterial oxygen saturation (Sa0₂) in a patient in whom pulsatile blood flow is reduced or absent, comprising:

alternately transilluminating a tissue of the patient with two or more wavelengths of light,

momentarily occluding arterial blood flow to the tissue of the patient, and then releasing occlusion of arterial blood flow to the tissue,

measuring change in transmission separately for each of the two or more wavelengths of light across the tissue during occlusion of arterial blood flow to the tissue and after release of occlusion of arterial blood flow to the tissue, where light levels at each wavelength at points of time after release of occlusion are expressed as a fraction of the maximum light level that occurs during occlusion, yielding a relative transmittance (T) level for each wavelength of light,

calculating arterial blood attenuance (A) for each wavelength at each time point, where A = log (1/T),

calculating an average arterial blood attenuance ratio for the two or more wavelengths, and

estimating arterial oxygen saturation (Sa0₂) in the patient by comparing the average arterial blood attenuance ratio with data from subjects with known Sa0₂ or with data from calibrating devices.

[0015] The invention also provides a system for estimating arterial oxygen saturation (Sa0₂) in a patient in whom pulsatile blood flow is reduced or absent, comprising:

one or more devices for emitting two or more wavelengths of light,

one or more light detecting devices, and

one or more computing devices comprising one or more processors, a memory unit, a display device, and a computer-readable storage medium including computer-readable code that is read by the one or more processors to perform a method comprising the steps of: alternately transilluminating a tissue of the patient with two or more wavelengths of light,

estimating arterial oxygen saturation (Sa02) in the patient by comparing the average arterial blood attenuance ratio with data from subjects with known SaC>2 or with data from calibrating devices.

[0016] The patient can, for example, have a peripheral vascular disease (PVD), circulatory shock, and/or be on a continuous flow left ventricular assist device (LVAD) or extracorporeal membrane oxygenator (ECMO). The patient can also have any other disease or condition, or be undergoing any procedure, where pulsatile blood flow is reduced or absent.

[0017] Suitable tissues for use in these procedures include, for example, fingers (e.g., fingertip), toes, ears (e.g., earlobe), and hands, although other tissues could also be used.

[0018] The arterial blood flow to the tissue can be occluded, e.g., for approximately 5- 10 seconds. Shorter or longer periods of occlusion are also suitable. In one embodiment venous blood flow from the tissue is also occluded during and for, e.g., approximately 2-5 seconds after arterial occlusion to the tissue. Venous occlusion can be applied, e.g., by a 20-30 mmHg cuff.

[0019] The two or more wavelengths of light used to transilluminate the tissue can include any suitable pair of wavelengths. Preferred examples include, but are not limited to, 660nm and 905nm, and 660 nm and 940 nm. One skilled in the art will recognize that any two wavelengths would be suitable, for which the relative absorption by oxy- and reduced hemoglobins differs substantially. Preferably, the tissue is alternately transilluminated by the two or more wavelengths of light at 100-700 Hz, for example at 480 Hz or 600 Hz. Preferably, the change in light transmission of each wavelength is measured before, during, and for at least 1-3 seconds after release of arterial occlusion. Shorter or longer periods of light transmission measurement are also suitable. Preferably, the transilluminated light is measured by a light-detecting device, such as a photodiode. Preferably, the output of the photodiode or other light measuring device is amplified and digitized at a resolution of 14 bit or higher.

[0020] In one embodiment, the average arterial blood attenuance ratio for two wavelengths is calculated by plotting arterial blood attenuance for one wavelength as a function of arterial blood attenuance for a second wavelength, where the slope of the plot represents the average arterial blood attenuance ratio for the two wavelengths. For example, a regression of arterial blood attenuance of 660nm wavelength (Red (R)) against arterial blood attenuance of 905nm wavelength (infrared (IR)) yields a slope estimate, equivalent to average R/IR ratio.

[0021] The measured R/IR ratio can be compared with a calibration equation or look-up table to yield an estimate of arterial oxygen saturation.

[0022] Calibrations for arterial oxygen saturation could also be obtained using measurements of R/IR ratios in subjects with known Sa0₂, at various levels of Sa0₂, achieved naturally or by exposing the subjects to hypoxic, normoxic, and hyperoxic gases. Alternatively, calibration could be achieved using a non-living calibrating device, such as an artificial finger.

[0023] One skilled in the art will recognize that ratios of changing transmission, rather than changing attenuance, compared with standards derived from subjects or nonliving calibrating devices with known Sa0₂, could be used to generate estimates of Sa0₂.

[0024] In the systems disclosed herein preferred devices for emitting light include, for example, two or more light emiting diodes (LEDs). Alternatively, one or more of the light emitting devices could be laser diodes, or incandescent or other light sources, modified by optical filters. Preferred light detecting devices include, e.g., at least one photodiode. Alternatively, a broad-band light source could be used, and optical filters could be interposed between the transilluminated finger or tissue and one or more light-detecting devices.

[0025] The system can include a device for occluding arterial blood flow to the tissue of the patient and/or a device for occluding venuous blood flow from the tissue of the patient. [0026] Apparatus and technique to transiently occlude and release the arteries supplying the relevant body part: Examples include occluding the radial and ulnar arteries, but other arteries could be used. Occlusion could be accomplished, e.g., by manual compression of the arteries; by automated compression, using a firm occluder device incorporated into a wristband or other device; by inflated cuff around the forearm, base of the fingertip, or other body part, or by other means. A second cuff inflated to a lower pressure might be used to prevent venous outflow upon release of occlusion.

[0027] Apparatus and technique to identify the adequacy of arterial occlusion during the occlusion/release maneuver are accomplished in one embodiment by monitoring the rate of rise in light level of one or more wavelengths, relative to baseline light level, and alerting the operator (or the software) if the rise or the rate of rise falls below a predetermined threshold.

[0028] Apparatus and technique to identify the adequacy of venous outflow during arterial occlusion to determine the optimal release point: This can be accomplished by monitoring the rate of rise in light level of one or more wavelengths and automating the release of occlusion when the rate of rise declines to a predetermined rate.

[0029] Examples of the apparatus include a pulse oximeter probe, containing two or more LEDs or other light-emitting devices, and at least one photodiode or other light- detecting device, interfaced with a microprocessor and a display. An occluder device can be incorporated, e.g., into a wrist-band, capable of applying momentary firm pressure to the ventral aspect of the wrist, specifically over the course of the radial and ulnar arteries, without impeding venous outflow from the hand. The device can be adjustable in position, to allow for varying patient wrist size and conformation.

[0030] The microprocessor can be programmed to detect the adequacy of pulsatile changes in light attenuance, and, if such changes were adequate, the device would function as a standard pulse oximeter, reporting pulse rate and pulse oximetric estimate of SaC>2 (Sp0₂).

[0031] If the microprocessor detects inadequate pulsations, it would activate an indicator on the display. When the operator desires an estimate of Sa0₂, the operator would provide input to the microprocessor, which would then initiate a series of several, e.g., 5 to 10 second occlusions of, e.g., the radial and ulnar arteries, monitored by the microprocessor as magnitude and rate of increase in light transmission at one or more wavelengths. When a predetermined increase in transmission has occurred, accompanied by a predetermined rate of decrease in transmission, or a predetermined time has passed, the arterial occlusion would be abruptly released. The microprocessor would perform the calculations illustrated in Figure 4 to determine, e.g., a R/IR ratio, and refer to a lookup table or regression equation to compute an estimate of SaC>2. A median result from, e.g., at least 3 occlusions could be reported as an estimate of Sa02, perhaps best abbreviated as SplC^ (for estimate of SaC>2 from pulseless oximetry).

[0032] An enhancement to the device would be the addition of estimates of carboxy- and met-hemoglobin levels, both during standard pulse oximetry and during the pulseless operation of the device. Carboxyhemoglobin and methemoglobin percentages would be calculated using multiple-regression-derived formulae including at least three variables: a ratio of changing attenuance of 630nm to changing attenuance at 905nm, a ratio of changing attenuance of 660nm to changing attenuance at 905nm, and a ratio of changing attenuance of 730nm to changing attenuance at 905nm, as previously described (Aldrich 2005). Many other pairs of wavelengths would be potentially suitable to make these calculations. One skilled in the art will recognize that many other ratios of changing absorbances at different wavelengths, selected as varying substantially in their absorbance of carboxy-hemoglobin, met-hemoglobin, oxyhemoglobin, and reduced hemoglobin, as described by Zjilstra (1991), could be used as independent variables in a formula to estimate each of the the various types of hemoglobin, separately.

[0033] This device would be particularly useful in vascular disease practices and heart failure programs, where pulselessness is not uncommon, and would be an important addition to operating rooms, endoscopy suites, emergency rooms, and general medical and surgical wards, which have to deal with pulseless patients.

[0034] The device could take the place of a standard pulse oximeter in many settings, functioning as a standard pulse oximeter, without intermittent arterial occlusion, and, because it could also measure methemoglobin (which can be seen in pulseless patients, e.g., vascular and heart failure patients, because of their exposure to nitrates) and carboxyhemoglobin, it would do a better job of screening and monitoring than a standard 2- wavelength pulse oximeter. The device would function as a standard pulse oximeter unless it detects pulselessness, in which case it would signal that fact and either automatically or when activated, shift to the pulseless mode and make its measurements.

[0035] This invention will be better understood from the Experimental Details, which follow. However, one skilled in the art will readily appreciate that the specific methods and results discussed are merely illustrative of the invention as described more fully in the claims that follow thereafter.

EXPERIMENTAL DETAILS

Introduction

[0036] This invention provides an approach to measuring arterial saturation noninvasively by comparing the change in transmittance (relative transmission) of two (or more) wavelengths observed across tissue (e.g., a finger, toe or hand) when arterial supply to the tissue is momentarily occluded or upon release of such occlusion. For practical purposes, the only change in transmittance observed upon release of occlusion is due to inflow of arterial blood, so comparison of pre- to post release attenuance (the logarithm of the reciprocal of transmittance) is proportional to arterial saturation.

Methods

[0037] Seven normal subjects were studied while they breathed room air, or hypoxic or hyperoxic gas mixtures. Four pulseless LVAD patients were studied (two on two occasions each) at their clinically-determined level of oxygenation.

[0038] Before, during and after approximately 5 second manual occlusions of radial and ulnar arteries (Figure 1), one fingertip was alternately trans illuminated by 2 light emiting diodes (LEDs), emitting approximately 660nm (red) and 905nm (infrared). The transilluminated light was monitored by a photodiode, with its output amplified and digitized with 16 bit resolution, and synchronized with the LED alternation to separately identify transmission of 660nm and 905nm. An example of the resulting data is shown in Figure 2-4. Data were collected in triplicate at each level of oxygenation studied.

[0039] For approximately 1 second after release of occlusion, raw light levels at each of the wavelengths at each time point (see Figure 3) were expressed as a fraction of the maximum for that particular wavelength, yielding relative transmittance (T) levels, demonstrating the decreasing light transmission caused by influx of arterial blood (see Figure 3, top two tracings).

[0040] For approximately 1 second after release of occlusion, arterial blood absorbance (actually attenuance, the sum of absorbance and scatter) was calculated for each wavelength at each time point as A=log(l/T) (see Figure 3, bottom two tracings). Arterial blood attenuance of 660nm was then plotted against arterial blood attenuance of 905nm, yielding a straight line (Figure 4), the slope of which represents the average red/infrared arterial blood attenuance ratio (R/IR).

[0041] Medians of the triplicate Sp0₂ results for each subject, at each level of oxygenation, were plotted as a function of median R/IR, yielding a preliminary calibration curve for the pulseless oximeter (see Figure 5).

Results

[0042] Five normal subjects were studied, four of them with multiple R/IR calculations at multiple SpC>2 levels obtained by breathing room air, or hypoxic or hyperoxic gas mixtures. No side effects of breathing the hyperoxic or hypoxic mixtures were noted. Seven pulseless LVAD patients were also studied, two at two levels of oxygenation each. R/IR calculations were obtainable in all experiments, with R² always > 0.98.

[0043] Figure 5 shows the calibration curve results. Medians of at least 2 (usually 3 to 5) replicate measurements are shown. The data follow an inverse relationship between SpC>2 and R/IR ratio. Data for the normal subjects (NS #1 through 5) defined a clear calibration curve. Data for the nine measurements in seven LVAD patients fell slightly above the calibration curve calculated for the normal subjects, underestimating SaC>2 by an average of 1.1 percentage points (maximum 3.4 percentage points), acceptable for clinical determinations of the adequacy of oxygenation.

Discussion

[0044] These results demonstrate that a comparison of changes in attenuance of light at two different wavelengths, transmitted through a fingertip, brought about by release of momentary arterial occlusion, could be used to estimate arterial saturation without reliance on pulsatile flow.

[0045] The experiments were performed with normal subjects at multiple levels of oxygenation and in seven patients with nonpulsatile left ventricular assist devices (LVADs). LVAD patients have normal or near normal cardiac output and normal or near normal blood flow into digits. They have no radial pulses, not because of impaired blood flow into the hands, but because their flow is nonpulsatile.

[0046] Patients with peripheral vascular disease (PVD) will have reduced blood flow into the hands, and so would be expected to show lesser amplitude of pbotoplethysmographic change after release of occlusion. However, (in the absence of gangrene), such patients do not have absent arterial blood flow and would be expected to sho photoplethysmographic changes upon relief of occlusion, changes that can be compared among wavelengths to estimate R/IR ratio. That being the case, even in PVD, the described technique is likely to be sufficient to differentiate normoxemic from hypoxemic patients.

[0047] Arterial occlusion allows venous blood to drain from the tissue of interest. The subsequent release of occlusion allows arterial blood to flow into the tissue, making, for practical purposes, arterial blood the only contributor to changing light transmission and attenuanee. Thus, unlike the pulse oximetry during artificial pulsations, the technique described in this application measures oxygenation of arterial blood, not venous blood or a mixture of arterial and venous blood. It is possible that continuing small changes in venous blood content of, e.g., the fingertip during occlusion and after release may contribute slightly to the observed changes in light transmission. For that reason, venous occlusion can be applied by, e.g., 20-30 mmHg cuff on, e.g., the forearm, during and for a few seconds after release of arterial occlusion.

[0048] Cuff-occlusion above arterial systolic blood pressure may be a suitable means of arterial occlusion, potentially simpler to automate than is manual occlusion. However, cuff occlusion would prevent venous outflow during the occlusion, preventing the rise in transmission that occurs during manual arterial occlusion, and may lead to more "contamination" of subsequent transmission measurements by changing venous blood content. That problem could be addressed by continuing post-release venous occlusion. So, for example, two cuffs could be applied to the forearm, one inflated to 30 mmHg and the other to ISOmmHg, both for 5 seconds. The 150mmHg cuff would be abruptly deflated to provide arterial blood flow to the tissue, and the 30 mmHg cuff would remain inflated for a further 2-5 seconds to maintain venous occlusion.

[0049] Unlike standard pulse oximetry, for which the pulsatile nature of the photoplethysmographic signals can be, and usually is, quantified as AC/DC ratios of normalized light level data (Palreddy 1997), the present technique cannot rely on measurements of the AC component of the photoplethysmographic signals. The present technique therefore requires discrimination of small, nonpulsatile light level changes. Digitization at 16 bit resolution was found to be adequate for that purpose.

[0050] In summary, the present invention provides a new approach to measure arterial oxygenation noninvasiveiy in patients in whom pulse oximetry fails: for example, those with poor peripheral pulses because of PVD or because they are on continuous flow left ventricular assist devices (LVADs). The technique involves a 5-second occlusion of, e.g., radial and ulnar arteries, followed by an abrupt release. A fingertip on the same hand is alternately transilluminated by two (or more) wavelengths of light, e.g. 660nm (red) and 905nm (infrared) at a rapid rate (~60()hz). Attenuance of each of the wavelengths is recorded during the 1-2 seconds after release of occlusion, and a ratio of changing attenuance of red to changing attenuance of infrared light (R/IR.) is calculated. R/IR can be converted to an estimate of arterial oxygen saturation (Sa0₂) by comparison with data from subjects with known SaO? or from data from a non-living calibrating device.

[0051] Experiments in five normal subjects at various Sa0₂ levels from 87% to 99% and in seven pulseless LVAD patients (two studied twice each) at Sa(3₂ 93% to 99% demonstrated an inverse relationship between R IR and SaO?_. Using data from the normals to generate a calibration curve, the error among LVAD patients averaged a clinically insignificant 1 percentage point, usually underestimated.

[0052] Additional ratios of changing attenuance of other pairs of wavelengths could be added for improved accuracy and, potentially, to allow for measurement of dyshemoglobins (eg carboxy- or met-hemoglobin) in addition to oxyhemoglobin.

[0053] As all the necessary components would be present for standard pulse oximetry, the device could be used for standard pulse oximetry and shift to the pulseless mode only when the pulse is inadequate for reliable pulse oximetry readings.

REFERENCES

Aldrich TK, Preiss D, Tso R. Measuring Carboxyhemoglobin and Methemoglobin Noninvasively. Proc Am Thorac Soc 2:A664, 2005. (Presented at the May 2005 American Thoracic Society meeting).

Palreddy, S. Signal processing algorithms. In: Design of Pulse Oximeters, edited by J. G. Webster. Bristol: IOP, 1997, pp. 124-58.

Weiben, O. Light absorbance in pulse oximetry. In: Design of Pulse Oximeters, edited by J. G. Webster. Bristol: IOP, 1997, pp. 40-55.

Zijlstra WGl, Buursma A, Meeuwsen-van der Roest WP. Absorption spectra of human fetal and adult oxyhemoglobin, de-oxyhemoglobin, carboxyhemoglobin, and methemoglobin. Clin Chem. 1991; 37: 1633-8.

Claims

What is claimed is:

1. A method of estimating arterial oxygen saturation (Sa0₂) in a patient in whom pulsatile blood flow is reduced or absent, comprising:

2. The method of claim 1, wherein the patient has peripheral vascular disease, circulatory shock, and/or is on a continuous flow left ventricular assist device (LVAD) or extracorporeal membrane oxygenator (ECMO).

3. The method of claim 1 or 2, wherein the tissue is a finger, toe or hand.

4. The method of any of claims 1-3, wherein the two wavelengths are 660nm and 905nm, or 660 nm and 940 nm.

5. The method of any of claims 1-4, wherein the tissue is alternately transilluminated by the two or more wavelengths of light at 100-700 Hz.

6. The method of any of claims 1-5, wherein trans illuminated light is digitized at 14 bit or higher resolution.

7. The method of any of claims 1-6, wherein the arterial blood flow to the tissue is occluded for 5-10 seconds.

8. The method of any of claims 1-7, wherein venous blood flow from the tissue is occluded during and for 2-5 seconds after arterial occlusion to the tissue.

9. The method of claim 8, wherein venous occlusion is applied by 20-30 mmlig cuff.

10. The method of any of claims 1-9, wherein change in light transmission of each wavelength is measured for 1-3 seconds after release of arterial occlusion.

11. The method of any of claims 1-10, wherein the average arterial blood attenuance ratio for two wavelengths is calculated by plotting arterial blood attenuance for one wavelength as a function of arterial blood attenuance for a second wavelength, where the slope of the plot represents the average arterial blood attenuance ratio for the two wavelengths.

12. The method of claim 11, wherein a regression of arterial blood attenuance of 660nm wavelength (Red (R)) against arterial blood attenuance of 905nm wavelength (infrared (IR)) yields a slope estimate, equivalent to average R/IR ratio.

13. The method of claim 12, wherein the measured R/IR ratio is compared with a calibration equation or look-up table to yield an estimate of arterial oxygen saturation.

14. A system for estimating arterial oxygen saturation (Sa02) in a patient in whom pulsatile blood flow is reduced or absent, comprising:

one or more devices for emitting two or more wavelengths of light,

one or more light detecting devices, and one or more computing devices comprising one or more processors, a memory unit, a display device, and a computer-readable storage medium including computer-readable code that is read by the one or more processors to perform a method comprising the steps of:

15. The system of claim 14 comprising a device for occluding arterial blood flow to the tissue of the patient.

16. The system of claim 14 or 15 comprising a device for occluding venuous blood flow from the tissue of the patient.

17. The system of any of claims 14-16, wherein the device for emitting light is two or more light emiting diodes (LEDs).

18. The system of any of claims 14-17, wherein the light detecting device is at least one photodiode.

19. The system of any of claims 14-18, wherein the two wavelengths are 660nm and 905nm, or 660 nm and 940 nm.

20. The system of any of claims 14-19, wherein the tissue is alternately trans illuminated by the two or more wavelengths of light at 100-700 Hz.

21. The system of any of claims 14-20, wherein trans illuminated light is digitized at 14 bit or higher resolution.

22. The system of any of claims 14-21, wherein change in light transmission of each wavelength is measured for 1-3 seconds after release of arterial occlusion.

23. The system of any of claims 14-22, wherein the average arterial blood attenuance ratio for two wavelengths is calculated by plotting arterial blood attenuance for one wavelength as a function of arterial blood attenuance for a second wavelength, where the slope of the plot represents the average arterial blood attenuance ratio for the two wavelengths.

24. The system of any of claim 23, wherein a regression of arterial blood attenuance of 660nm wavelength (Red (R)) against arterial blood attenuance of 905nm wavelength (infrared (IR)) yields a slope estimate, equivalent to average R/IR ratio.

25. The system of claim 24, wherein the measured R/IR ratio is compared with a calibration equation or look-up table to yield an estimate of arterial oxygen saturation.