WO2002080764A9

WO2002080764A9 - Non-invasive determination of blood components

Info

Publication number: WO2002080764A9
Application number: PCT/US2002/010804
Authority: WO
Inventors: Gregory D Jay; Selim Suner; Kevin S Ernsting
Original assignee: Rhode Island Hosp Lifespan Ptr; Gregory D Jay; Selim Suner; Kevin S Ernsting
Priority date: 2001-04-05
Filing date: 2002-04-05
Publication date: 2003-01-03
Also published as: AU2002309545B2; EP1383422A1; US20030002722A1; US7711403B2; CA2442921A1; WO2002080764A1; JP4467238B2; EP1383422A4; JP2004535219A

Abstract

A method of estimating an amount of a substance in a bodily fluid from the color of a tissue surface of a subject (20) includes capturing an image including at least a portion-of-interest of the subject and at least a reference portion of a color reference, the image being a digital image of pixels of at least one color, the image including a component value for each pixel for each of the at least one color, obtaining a first value associated with at least one component value corresponding to the portion-of-interest of the subject, obtaining a second value associated with at least one component value corresponding to the reference portion, and calculating an estimated amount of the substance using the first and second values.

Description

NON-INVASIVE DETERMINATION OF BLOOD COMPONENTS

FIELD OF THE INVENTION

The invention relates to determining hemoglobin concentrations and more particularly

to determining hemoglobin concentrations in a non-invasive manner.

BACKGROUND OF THE INVENTION

Determining hemoglobin (Hb) concentrations is often useful in diagnosis and treatment

of patients. For example, Hb concentrations are useful in diagnosing whether a person is

anemic. Several techniques currently exist for determining serum hemoglobin. For example,

techniques include phlebotomy (puncturing a vein and drawing blood) with subsequent

laboratory analysis of the drawn blood, microscopic assessment of mucous membranes, and

subjective assessment of conjunctiva color and clinical parameters. By far the most widely

used technique is phlebotomy followed by analysis with a flow cytometer. This technique is

reliable, but uses one or more technicians, an expensive piece of equipment (the cytometer),

and typically a centralized location for the analysis due to the cost of the cytometer. Another

technique uses magnification and analysis of blood flow through mucous membranes.

SUMMARY OF THE INVENTION

In general, in an aspect, the invention provides a method of estimating an amount of a

substance in a bodily fluid from the color of a tissue surface of a subject. The method includes

capturing an image including at least a portion-of-interest of the subject and at least a reference

portion of a color reference, the image being a digital image of pixels of at least one color, the

image including a component value for each pixel for each of the at least one color, obtaining a first value associated with at least one component value corresponding to the portion-of-

interest of the subject, obtaining a second value associated with at least one component value

corresponding to the reference portion, and calculating an estimated amount of the substance

using the first and second values.

Implementations of the invention may include one or more of the following features.

The image comprises a plurality of colors for each pixel, wherein obtaining the first value

comprises obtaining a plurality of first values each associated with at least one component

value of a corresponding color, wherein obtaining the second value comprises obtaining a

plurality of second values each associated with at least one component value of a

corresponding color, and wherein the calculating the estimated amount uses each of the

plurality of first and second values. The obtaining the first and second values comprises

averaging component values in a first portion of the portion-of-interest and a second portion of

the reference portion, respectively, for each of the plurality of colors. The calculating

comprises using first and second empirically-predetermined weighting constants associated

with the first and second values respectively.

Implementations of the invention may also include one or more of the following

features. The method further includes selecting a first portion of the portion-of-interest and a

second portion of the color reference, and wherein the first and second values are obtained

from at least one component value of each of the first and second portions, respectively. The

first portion is a biologically perfused surface of the subject. For example, a biologically

perfused surface or tissue is one which contains blood vessels. Preferably, the tissue or surface

is richly vascularized, e.g., the conjunctiva of the eye or soft tissues such as a mucous

membrane. More preferably, the vascularized tissue is exposed on the surface of the body of

an individual, e.g., skin or nail bed, or easily accessible, e.g., the conjunctiva of the eye or a mucous membrane. The capturing is performed non-invasively.

Implementations of the invention may also include one or more of the following

features. The capturing, the obtaining a first value, the obtaining a second value, and the

calculating are performed in real time. The obtaining a first value, the obtaining a second

value, and the calculating are at least partially performed by a processor executing software

instructions. The portion-of-interest is a conjunctiva and the amount of the substance is a

hemoglobin concentration. The method further includes displaying the estimated amount. The

estimated amount is a hemoglobin concentration and is displayed in units of grams per

deciliter (g/dl).

In general, in another aspect, the invention provides a method of determining factors

that influence estimation of an amount of a substance in a bodily fluid from the color of a

tissue surface of a subject. The method includes capturing an image including at least a

portion-of-interest of the subject and at least a reference portion of a color reference, the image

being a digital image of pixels of at least one color, the image including a component value for

each pixel for each of the at least one color, establishing a first factor associated with the

portion-of-interest and a second factor associated with the reference portion, calculating an

estimated amount of the substance in the subject using the component values and the first and

second factors, comparing the estimated amount with a corresponding known amount of the

substance in the subject, and adjusting, if the estimated amount has other than a desired

relationship with the known amount, at least one of the first and second factors.

The method further includes repeating the calculating, the comparing, and the adjusting until

the estimated amount of the substance has the desired relationship with the known amount of the substance. The method further includes storing the first and second factors when the estimated amount of the substance has the desired relationship with the known amount of the

substance. The image comprises a plurality of colors for each pixel, wherein the establishing

establishes a plurality of first factors and a plurality of second factors each corresponding to a

respective one of the plurality of colors, wherein the calculating the estimated amount uses

each of the first and second factors, and wherein the adjusting adjusts any of the first and

second factors. The method further includes selecting a first region of the portion-of-interest

and a second region of the reference color. The calculating uses averages of the component

values for each color over the first and second regions, respectively. The calculating uses

ratios of the averages of the component values for each color over the first and second regions,

respectively.

In general, in another aspect, the invention provides a system for determining a level of

a substance in a bodily fluid from the color of a tissue surface. The system includes a color

separator module configured to decompose a digital color image of a perfused surface of a

subject and a color reference object into sub-images of component colors, the sub-images

comprising digital component values corresponding to pixels of the image, a portion selector

module in communication with the color separator module and configured to select a first

window of the image of the perfused surface and to select a second window of the image of the

color reference object, and a substance estimator module in communication with the portion

selector and configured to calculate an estimated level of the substance using window values

associated with component values corresponding to the first and second windows.

The system further includes an imaging module configured to capture the image and to convey

the digital component values of the image to the color separator module. The imaging module

is configured to capture the image non-invasively. The portion selector module is further configured to average the component values of each sub-image to produce the window values,

with one window value for each sub image for each of the first and second windows. The

color separator module, the portion selector module, and the substance estimator module each

comprise computer-executable instructions, stored on a computer-readable medium, for

causing a computer to perform actions as recited in claim 20. The system further includes a

display coupled to the substance estimator module and configured to display indicia of the

estimated level of the substance.

Various aspects of the invention may provide one or more of the following advantages. Hemoglobin concentrations are determined regardless of lighting conditions, in real-time, at

remote locations, and with inexpensive equipment. Hemoglobin concentrations are

determined objectively, reliably, and repeatably. Hemoglobin concentrations are determined

using, e.g., a portable digital camera including specialized software. Hemoglobin

concentrations are determined non-invasively and using widely available resources.

These and other advantages of the invention, along with the invention itself, will be

more fully understood after a review of the following figures, detailed description, and claims.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a schematic perspective view of a system for estimating biologically-relevant factors.

FIG. 2 is a block diagram of modules of a digital camera shown in FIG. 1.

FIG. 3 is a simplified diagram of an image, taken using the camera shown in FIG. 1,

including a portion of a subject and a color-reference card.

FIG. 4 is a block flow diagram of a process of estimating hemoglobin concentration.

FIG. 5 is a block flow diagram of a process of determining weighting constants for estimating hemoglobin concentration.

FIG. 6 is a block flow diagram of a process of estimating hemoglobin concentration

using the weighting constants determined using the process shown in FIG. 5.

FIG. 7 is a scatter plot of experimental data showing predicted hemoglobin

concentration, using the process shown in FIG. 5, and actual measured hemoglobin

concentration. Hemoglobin is expressed as grams per deciliter (g/dL).

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

The invention provides techniques for determining the concentration of a substance in

bodily fluids such as blood for diagnostic purposes. For example, the methods are useful to

determine the concentration of hemoglobin, methemoglobin, carboxyhemoglobin, bilirubin,

and bile salts to determine whether an individual is suffering from or at risk of developing a

pathological condition such as anemia, dyshemoglobinemia, methemoglobinemia,

carboxyhemoglobinemia, and carbon monoxide poisoning. Bilirubin disorders such as

hyperbilirubinemia and bilirubinuria are also diagnosed using the methods described herein.

In accordance with the invention, a user digitally photographs a patient and a gray

reference simultaneously and adjacently. The digitally-captured image is processed based on known relationships of images and the gray reference with known hemoglobin concentrations

or concentrations of other blood components. From the processing, a hemoglobin

concentration figure is produced. This figure can be used to diagnose hemoglobin-related

health issues (or at least health issues for which hemoglobin concentrations are indicated). For

example, hemoglobin levels outside a normal range (14-18 g/dL for human male subjects; 12-

16 g/dL for human female subjects) or bilirubin concentrations outside a normal range (0.3-1.0

mg/dL) indicate a pathological condition or a risk of developing such a condition. Referring to FIG. 1, a system 10 includes a subject 12, a gray-reference card 14, a

digital camera 16, and a user 18. The subject is a person whose hemoglobin concentration is

to be determined. The gray-reference card 14 is a known shade of gray, here an 18%

photographic standard gray card (i.e., the card is 18% of the way from pure white to pure

black). The card 14 is disposed adjacent an area 20 of the subject 12 to be imaged. Preferably,

the area 20 is a biologically perfused surface that's color is influenced by the subject's blood.

For example, the area 20 can be, e.g., the subject's conjunctiva, a mucous membrane, nail bed,

earlobe, skin, lips, or retina. The camera 16 is a digital camera configured to be manipulated

by the user 18 to obtain a digital image of the area 20 of the subject 12. The camera 16 images

the area 20 as a set of pixels with red, green, and blue components with corresponding red,

green, and blue values for each pixel. Here, the camera 16 can resolve the level of red, green,

and/or blue for any given pixel into one of 256 different shades, yielding a composite image of

over 16 million colors (256*256*256). The camera 16 is configured with computer-

executable software to process levels of red, green, and blue of the area 20 as compared to the

card 14 as described below to determine a hemoglobin concentration of the subject 12. The

camera 16 includes a display screen for displaying estimated hemoglobin concentrations in

units of grams per deciliter (g/dl).

Referring also to FIG. 2, the digital camera 16 includes an imaging module 21, an

outlining software module 22, a color separation software module 24, and a hemoglobin

concentration estimator software module 26. Each of the modules 22, 24, 26 includes

computer-executable instructions stored on a computer-readable medium such that the

instructions can be read and executed by a processor 32 of the camera 16. The modules may be referred to as performing actions as shorthand for the processor 32 executing the modules'

instructions for performing the listed actions/functions. The imaging module 21 includes both hardware and software. The imaging module 21 is configured to cause the camera 16 to take

pictures and capture color images, the images being composed of pixels of red, blue, and green

components having respective red, blue, and green component digital values.

The outlining module 22 is configured to select portions of an image corresponding to

the area 20 of the subject 12, and the card 14. As shown in FIG. 3, the outlining module can

select or outline a window 34 in the region 20 of the subject 12, and a window 36 of the card

14, from an image 38 taken by the camera 16. As shown, the region 20 is the conjunctiva, of

the subject 12, that has been exposed by turning down the lower eyelid of the subject 12.

The color separation module 24 is configured to use imaging software, e.g., available

from the National Institute of Health (NTH) (NTH Image for Macintosh computers and Scion

Image for PCs; information is available at http ://rsb .info .nih.bov/nih-image . The module 24 uses the imaging software to separate or deconstruct the colors imaged by the camera 16 into

the components of the pixels forming the image 38. Thus, three sub-images similar to those

shown in FIG. 3 are produced, one in red, one in green, and one in blue. The module 24 is

further configured to average the digital component values over the outlined regions 34, 36 and

to determine ratios of average component values for red, green, and blue of the window 34 of

the subject 12 and of the window 36 of the gray-reference card 14.

The estimator module 26 is configured to use ratios of values of the separated colors

relative to the card 14 from the color separation module 24, along with predetermined

weighting constants, as inputs to formulas that produce a hemoglobin estimate. The estimator module 26 is configured to select measured values from the imaging module 24 corresponding

to different selected portions of the image 38 taken by the camera 16 and to use the selected

measured values to determine the hemoglobin concentration estimate. Specifically, the

estimator module 26 is configured to determine a hemoglobin estimate Hb_est according to: Hb_est = 1.15 * ((hgb +hgbl)/2) (1)

where

hgbl = (120 - ((bme_c - gme_c) + (120 - (gme_c - rme_c) - 20) - 20))/12 (2)

and

hgb = (40 + ((160 - (((bme_c - gme_c) + (140 - (gme_c - rme_c)) - 12)))/4))/4.5 (3)

bmult = 150 bme_s (4)

gmult = 165/gme_s (5)

rmult = 0.8 + (130/rme_s) (6)

where bme_s, gme_s, and rme_s are the means (averages) of the blue, green, and red,

respectively, component values for the pixels in the window 36 of the reference card 14 used

as a standard, and where bme_c, gme_c, and rme_c are the means (averages) of the blue, green, and

red, respectively, component values for the pixels in the window 34 of the area 20, here the

subject's conjunctiva, multiplied by bmult, gmult, and rmult, respectively. Thus, bme_c, gme_c,

and rme_c include ratios of the averages of the component values in the window 34 and the

averages of the component values in the window 36. The constants, i.e., 150 in equation (4),

165 in equation (5), and 0.8 and 130 in equation (6) are weighting constants empirically

determined as described below.

In operation, referring to FIG. 4, with further reference to FIGS. 1-3, a process 50 for

estimating the subject's hemoglobin concentration using the camera 16 includes the stages shown. The process 50, however, is exemplary only and not limiting. The process 50 can be

altered, e.g., by having stages added, removed, or rearranged. The process 50 includes a stage 52 where weighting constants are determined, and a stage 54 where the determined weighting constants are applied to an image of a subject to estimate the subject's hemoglobin.

Referring also to FIG. 5, stage 52 of the process 50 includes the stages shown for

determining the weighting constants to be used at stage 54. The stages shown are exemplary

only and not limiting. Stages may be added, removed, or rearranged.

At stage 60, the image 38 of the subject 12 and the card 14 is taken using the camera

16. The card 14 is placed adjacent the region 20 to be imaged, here the subject's conjunctiva.

The subject 12 everts the subject's lower eyelid and holds the card 14 adjacent to the subject's

head at eye level. The user 18 actuates the camera 16 to take the image 38.

At stage 62, the windows 34, 36 of the subject 12 and the card 14 are outlined. The

outline module 22 outlines the windows 34, 36 in the region 20 of the subject 12 and within

the perimeter of the card 14

At stage 64, the image 38 is separated into red, green, and blue images. The color

separation module 24 separates the pixels of the image 38 using the imaging software (e.g.,

NTH Image or Scion Image). The module 24 further determines ratios of red, green, and blue

values to a value associated with the gray-reference card.

At stage 66, a weighting constant is determined for each color (red, green, and blue)

using known hemoglobin concentrations. Values of ratios from stages 62, 64, and 66 for

several subjects 12 are provided, along with a known hemoglobin concentrations of these same

subjects 12 (e.g., derived from phlebotomy and laboratory analysis). The provided values are

entered into equations (l)-(6) with initial red, green, and blue weighting constants. The initial

constants are arbitrary, e.g., 1, but may be selected to help reduce the number of iterations

involved in determining final weighting constants.

At stage 68, the subject's hemoglobin concentration is estimated by the concentration

estimator module 26. The module 26 uses the values of the ratios from stage 66 in equations (l)-(6) to estimate the subject's hemoglobin concentration.

At stage 70, the estimated and known concentrations are compared. The estimated

versus known hemoglobin concentrations are compared to determine whether adjustments to

one or more of the weighting constants is appropriate (i.e., if the estimated and known

concentrations meet or do not meet desired criteria such as being within a desired percentage

of each other).

At stage 72, appropriate adjustments are made to the red, green, and/or blue weighting

constants. The weighting constants are adjusted to attempt to achieve a more accurate

hemoglobin concentration estimate. If any adjustment is made to a weighting constant, then

the process 60 returns to stage 68 where the hemoglobin concentration for a particular subject

12 is determined. If no adjustments are made, then the process 60 proceeds to stage 74.

At stage 74, the weighting constants are stored for future use. The red, green, and blue

weighting constants are stored for use in determining estimates of hemoglobin concentrations

for subjects 12 whose hemoglobin may not be determined through other means, e.g.,

phlebotomy and lab analysis.

Stage 52 thus provides stage 54 with red, green, and blue weighting constants for use in

estimating hemoglobin concentrations for other subjects 12. Referring also to FIG. 6, stage 54

of the process 50 includes the stages shown for using the determined the weighting constants

to estimate the subject's hemoglobin concentration. The stages shown are exemplary only and

not limiting. Stages may be added, removed, or rearranged.

At stage 80, stages 60, 62, and 64 shown in FIG. 5 are performed for a non-reference

subject 12 whose hemoglobin concentration is to be estimated. Typically, this subject 12 will not have hemoglobin concentration determined through other means. The subject 12 may be located far away from other equipment needed for actual determination or direct measurement of hemoglobin concentration.

At stage 82, the ratios of color values to the gray card 14 determined at stage 80 are

applied to estimate the subject's hemoglobin concentration. The concentration estimator

module 26 uses the ratios provided from the color separation module 24 at stage 80 and the

weighting constants determined in stage 52 in equations (l)-(3) to determine an estimation of

the subject's hemoglobin concentration. The module 26 provides a number indicating the

hemoglobin concentration upon which a diagnosis of the subject 12 may be based.

Stage 52 will typically occur well before stage 54, although this is not necessary. For

example, the image 38 of the subject 12 may be taken at stage 80, then the weighting constant

determined at stage 52, then the subject's hemoglobin estimated at stage 82. Further, after a

hemoglobin concentration is estimated, weighting constants may be updated/refined and

hemoglobin estimates recalculated.

Referring to FIG. 7, experimental results using the process 50 evidenced a correlation

between hemoglobin concentration and conjunctiva color. For the experiments, persons with

active bleeding, oxygen saturation less than 90%, or a serum bilirubin over 3.0 mg/dL were

excluded. Actual Hb was measured with a cell counter model SE 9500 made by Sysmex Corp.

of Kobe, Japan. Photographs were taken with a digital camera model DSC-F1, with 480 x 260

pixels, made by Sony Corp. of Tokyo, Japan. 117 images were used from 63 patients (79

images for formula derivation and 38 for a validation group). 46% of the patients were female,

and the patients ranged in age from 20 to 87 years, with a mean 60 + 18. The formula

derivation group had a Pearson's Rank Order Coefficient of r(77) = 0.634, p<0.01. The

validation group had a Pearson's Rank Order Coefficient of r(36)=0.522, pθ.01. The

standard error was 2.57 and the standard deviation was 3.09.

Other embodiments are within the scope and spirit of the appended claims. For example, due to the nature of software, functions described above can be implemented using

software, hardware, firmware, hardwiring, or combinations of any of these. Features

implementing functions may also be physically located at various positions, including being

distributed such that portions of functions are implemented at different physical locations. For

example, portions of the software described above as being in the camera 16 can be disposed

externally to the camera 16. One or more of the software modules may be disposed in an

external computer, such as a laptop computer, or on a computer-readable medium such as a

floppy disc or compact disc (including a re-writable compact disc). In these cases, images

taken by the camera 16 can be loaded onto a computer that executes the software externally to

the camera 16. The computer or other external device can display estimated amounts in

appropriate units, such as hemoglobin concentrations in units of grams per deciliter (g/dl).

What is claimed is:

Claims

1. A method of estimating an amount of a substance in a bodily fluid from the

color of a tissue surface of a subject, the method comprising:

capturing an image including at least a portion-of-interest of the subject and at least a

reference portion of a color reference, the image being a digital image of pixels of at least one

color, the image including a component value for each pixel for each of the at least one color;

obtaining a first value associated with at least one component value corresponding to

the portion-of-interest of the subject;

obtaining a second value associated with at least one component value corresponding

to the reference portion; and

calculating an estimated amount of the substance using the first and second values.

2. The method of claim 1 wherein the image comprises a plurality of colors for

each pixel, wherein obtaining the first value comprises obtaining a plurality of first values each

associated with at least one component value of a corresponding color, wherein obtaining the

second value comprises obtaining a plurality of second values each associated with at least one

component value of a corresponding color, and wherein the calculating the estimated amount

uses each of the plurality of first and second values.

3. The method of claim 2 wherein obtaining the first and second values comprises

the reference portion, respectively, for each of the plurality of colors.

4. The method of claim 3 wherein the calculating comprises using first and second

empirically-predetermined weighting constants associated with the first and second values

respectively.

5. The method of claim 1 further comprising selecting a first portion of the

portion-of-interest and a second portion of the color reference, and wherein the first and

second values are obtained from at least one component value of each of the first and second

portions, respectively.

6. The method of claim 5 wherein the first portion is a biologically perfused

surface of the subject.

7. The method of claim 6 wherein the capturing is performed non-invasively.

8. The method of claim 1 wherein the capturing, the obtaining a first value, the

obtaining a second value, and the calculating are performed in real time.

9. The method of claim 1 wherein the obtaining a first value, the obtaining a

second value, and the calculating are at least partially performed by a processor executing software instructions.

10. The method of claim 1 wherein the portion-of-interest is a conjunctiva and the

amount of the substance is a hemoglobin concentration.

11. The method of claim 1 , wherein the substance is selected from the group consisting of hemoglobin, methemoglobin, carboxyhemoglobin, bilirubin, and a bile salt.

12. The method of claim 1 further comprising displaying the estimated amount.

13. The method of claim 12 wherein the estimated amount is a hemoglobin

concentration and is displayed in units of grams per deciliter (g/dl).

14. A method of determining factors that influence estimation of an amount of a

substance in a bodily fluid from the color of a tissue surface of a subject, the method

comprising:

establishing a first factor associated with the portion-of-interest and a second factor

associated with the reference portion;

calculating an estimated amount of the substance in the subject using the component

values and the first and second factors;

comparing the estimated amount with a corresponding known amount of the substance

in the subject; and adjusting, if the estimated amount has other than a desired relationship with the known

amount, at least one of the first and second factors.

15. The method of claim 14 further comprising repeating the calculating, the

comparing, and the adjusting until the estimated amount of the substance has the desired relationship with the known amount of the substance.

16. The method of claim 15 further comprising storing the first and second factors

when the estimated amount of the substance has the desired relationship with the known

amount of the substance.

17. The method of claim 14 wherein the image comprises a plurality of colors for

each pixel, wherein the establishing establishes a plurality of first factors and a plurality of

second factors each corresponding to a respective one of the plurality of colors, wherein the

calculating the estimated amount uses each of the first and second factors, and wherein the

adjusting adjusts any of the first and second factors.

18. The method of claim 14 further comprising selecting a first region of the

portion-of-interest and a second region of the reference color.

19. The method of claim 18 wherein the calculating uses averages of the

component values for each color over the first and second regions, respectively.

20. The method of claim 19 wherein the calculating uses ratios of the averages of

the component values for each color over the first and second regions, respectively.

21. A system for deteπnining a level of a substance in a bodily fluid from the color of a tissue surface, the system comprising:

a color separator module configured to decompose a digital color image of a perfused surface of a subject and a color reference object into sub-images of component colors, the sub-

images comprising digital component values corresponding to pixels of the image;

a portion selector module in communication with the color separator module and

configured to select a first window of the image of the perfused surface and to select a second

window of the image of the color reference object; and

a substance estimator module in communication with the portion selector and

configured to calculate an estimated level of the substance using window values associated

with component values corresponding to the first and second windows.

22. The system of claim 21 further comprising an imaging module configured to

capture the image and to convey the digital component values of the image to the color

separator module.

23. The system of claim 22 wherein the imaging module is configured to capture

the image non-invasively.

24. The system of claim 21 wherein the portion selector module is further

configured to average the component values of each sut image to produce the window values,

with one window value for each sub image for each of the first and second windows.

25. The system of claim 21 wherein the color separator module, the portion selector

module, and the substance estimator module each comprise computer-executable instructions,

stored on a computer-readable medium, for causing a computer to perform actions as recited in

' claim 20.

26. The system of claim 21 further comprising a display coupled to the substance

estimator module and configured to display indicia of the estimated level of the substance.

27. The system of claim 21, wherein the substance is selected from the group

consisting of hemoglobin, methemoglobin, carboxyhemoglobin, bilirubin, and a bile salt.