January
2005, Issue 174
Light-to-Frequency
Conversion (Part 2)
Pulse
and Oxygen Content
O2
SATURATION
Before
the use of pulse oximetry, the only way to monitor oxygen
levels in blood was to take a blood sample. Discounting
the invasiveness of the procedure, this required getting
a laboratory involved. Needless to say, this wasn’t
a real-time process by any means.
My
four-month-old grandson Joshua recently had surgery
(see Photo 1). As my family huddled around him in the
recovery room, I asked one of the nurses about the monitor
they were using to display bpm and O2. I have the sneaking
suspicion she doesn’t get asked technological questions
too often because her eyes popped open and she immediately
went into a sales pitch about how this tool improves
patient care.
|
(Click
here to enlarge)
|
Photo
1—Joshua rejoices in having the nose tube and oxygen
mask removed after surgery. The inset shows a pulse
oximetry sensor applied to his big toe. The glow
of a red LED triggered the technical questions I
put to the nurse on duty. |
Calculating
O2 saturation is based on the fact that the light absorption
is dependent on both the wavelength and material. Deoxygenated
blood absorbs red light (600 to 700 nm) at a greater
rate than IR light (800 to 940 nm). This means that
by comparing the measurements made at the two wavelengths,
you can calculate the amount of oxygen in the blood.
The maximum and minimum excursion values (AC portion)
of the sampled data are related to the mean value (DC
portion). So, it’s the ratio of these values that can
be compared between each of the two wavelengths. Depending
on which LED is on, either IR_Ratio or RED_Ratio is
calculated once per heartbeat.
These
ratios are used to calculate O2 saturation when PB4
has requested O2 mode. RED_Ratio divided by IR_Ratio
should provide a value less than 10. This nonlinear
value relates to an oxygen level where 0.4 corresponds
to 100%, 1 corresponds to 85%, and 3.4 corresponds to
0%. Here I’m interested in values between 80% and 100%.
For this project, I assumed this portion of the curve
to be linear.
There
are plenty of opportunities for error when trying to
determine O2 saturation. The most obvious sources of
error are the actual LED wavelengths of the devices.
Extraneous light, sensor movement, light-absorbing species
in the blood (e.g., dyes and gases like carbon monoxide),
and poor blood flow also can causes errors. A tool is
only as good as the technician working with it.
