Our goal is to quantify muscle activity and fatigue using LED spectroscopy to track oxygen trends in muscle tissue over time. Near-infrared light penetrates tissue better than visible light and can detect deoxygenated hemoglobin and myoglobin in muscle. Using a single infrared LED, we can track relative oxygen saturation trends but cannot account for changing path length within muscle during movement. To normalize signals across subjects, we measure maximal deoxygenation during induced ischemia. Infrared signal declines seen in most samples may indicate decreased blood flow or oxygen consumption exceeding supply during static loads. Moving forward, combining ultrasound and infrared signals could help separate muscle influence from oxygen changes and classify activity.

1. Pulse Oximeter Sensor

2. Our Goal
We hope to quantify the dynamics of an active
muscle.
LED spectroscopy provides a way to track fatigue
and oxygen trends of an area of interest.

6. Properties of the Red and IR LED
• Near-IR light has much better penetration into
biological tissue than visible light
• NIR light absorption affected by: hemoglobin,
myoglobin and cytochrome c oxidase molecules
(90%Hb)
• To track specifically muscle activity, it appears
that Infrared light will be the most appropriate
signal to track

7. IR signal
We know from Beer-Lambert’s law that the absorption of
an LED is a function of each absorbed substance given
their:
1. extinction factor
2. concentration
3. path length.
Our IR light detects mostly Hb/Mb-deoxygenation in
muscle tissue but we cannot account for path length
using 1 LED.

8. Path Length
For a fixed optode (forehead) we could assume
path length to be constant.
However, we are interested in quantifying
muscle movement within exercise, and our
signal includes an initial contraction of muscle
(change of path length)

9. Accounting for path length
The optical path length was accounted for by
measuring the absorbance changes obtained during
cuff ischemia.
The percent deoxygenation during exercise is then
calculated from the maximal deoxygenation of the
ischemic response.
A way of normalizing NIRS signals across subjects.

10. Oxygen saturations
Boushel claims one must use multiple
wavelengths to generate algorithms for NIRS.
• Coefficients used at each wavelength
deconvolute the overlapping absorption
spectra.
• Algorithms are used for either trend
monitoring or to assess relative
concentrations.

11. Below are wavelengths and coefficients used in
the laboratory at Duke for assessing relative
concentrations.

12. Measurements with Two
Wavelengths
Other NIRS applications use two near-IR LEDs
(760 and 850nm) and use the difference
between the signals to determine relative
oxygen saturation. The sum of the signals to
provide a reference to blood volume changes.

13. Red-IR= relative O2
While I am using a
Red LED and IR LED,
the data shows
plausible oxygen
trends in that the
bicep oxygen content
desaturates with
increased load in
isometric exercise.

14. Infrared Trends
We changed the experiment set up
Infrared signals have shown a promising fatigue
trend in almost all of our samples.

16. Analyzing IR trends
Static load in some types of activity compromise
blood flow to working muscle.
Are our IR decline trends due to decreased
blood flow, or is the use of oxygen in the muscle
exceeding a continuous supply?

17. Moving Forward
Ultrasound & IR signal of Nellcor sensor in
cooperation. Separate muscle influence and oxygen
Classify experiment activity, static load or active
Use multiple NIRS LEDs?
• Overlapping spectra= trend monitoring/ find
relative concentrations of each molecule.