Pulse Oximeter

Hi All,

I would like to make an electronic “finger” so to speak that will be able to use two LED’s at the tip (one Red 660nm and one IR 940nm) this is to project IR and Red light signals that can be read by a standard off the shelf pulse oximeter.

For those that are not familiar with pulse oximetry here is a link with some information:

http://www.freescale.com/files/32bit/doc/app_note/AN4327.pdf?tid=AMdlDR

The problem is in normal pulse oximetry the Hemoglobin acts as filter that absorbs some of the incoming light. I need to make a circuit that simulates this filtration varying the light being sent out of the finger to the pulse oximeters light sensor.

I have attached a CRUDE drawing of what I am trying to accomplish.

Thanks in advance

I would first attempt to do this a bit more mechanically. Find a pair of filters which simulate the pulse states and attach them to the arm of a servo. Use the servo to drive the filters into the sensor.

It’s not entirely clear from your drawing but it looks like you have blocked the oximeter’s own LEDs and are substituting your own. This won’t work as the LED are pulsed at relatively high frequency. This is necessary to eliminate the effect of ambient IR light (from the sun.) The sensor is looking at the difference between when its own light is on versus off and it might be doing this at more than 1000Hz. It may be possible by putting your own light sensor under the two oximeter LEDs and then synchronising to that but the timing requirements are going to be extremely tight.

In fact, I rather suspect that the two LEDs in the oximeter probe are connected back-to-back and powered alternately so that the (single) detector assesses the two colours alternately (and a reference with neither illuminated).

(Pretty sure I have one around here somewhere that I managed to souvenir when I got out of ICU ...)

MorganS: I would first attempt to do this a bit more mechanically. Find a pair of filters which simulate the pulse states and attach them to the arm of a servo. Use the servo to drive the filters into the sensor.

It's not entirely clear from your drawing but it looks like you have blocked the oximeter's own LEDs and are substituting your own. This won't work as the LED are pulsed at relatively high frequency. This is necessary to eliminate the effect of ambient IR light (from the sun.) The sensor is looking at the difference between when its own light is on versus off and it might be doing this at more than 1000Hz. It may be possible by putting your own light sensor under the two oximeter LEDs and then synchronising to that but the timing requirements are going to be extremely tight.

My only issue with doing it mechanically with a servo is that all of the components need to fit within a finger. My goal is to be able to trick an off the shelf pulse oximeter into thinking a real finger is inside of it.

I was thinking I would use a Red LED and IR LED inside of the finger and pulse those LED's at a certain frequency that would then be read by the sensor that is at the bottom of the pulse Ox

zjjackson101: I was thinking I would use a Red LED and IR LED inside of the finger and pulse those LED's at a certain frequency that would then be read by the sensor that is at the bottom of the pulse Ox

See my reply above!

Paul__B: In fact, I rather suspect that the two LEDs in the oximeter probe are connected back-to-back and powered alternately so that the (single) detector assesses the two colours alternately (and a reference with neither illuminated).

(Pretty sure I have one around here somewhere that I managed to souvenir when I got out of ICU ...)

Yes I believe that is how they work.

Any Idea how I can Pulse My LED's to imitate the ones in the Pulse Oximeter?

My goal is to be able to trick an off the shelf pulse oximeter into thinking a real finger is inside of it.

Why? Is it to bypass some security system?

Grumpy_Mike: Why? Is it to bypass some security system?

Haha no,

The long term goal is to create a full healthcare simulator, Im starting with this and a few other small pieces to the project though

From what I have read about pulse oximetry, I am having a hard time understanding if i should make a circuit that varies the pulse frequency of the RED and Near IR LED's. Or if I should make a circuit that varies the intensity of the LED's

zjjackson101: Any Idea how I can Pulse My LED's to imitate the ones in the Pulse Oximeter?

So you want to create a "dummy finger" eh?

Well (given the detector operates as I say, two LEDs and one photoidiode or photo-transistor), you need a pair of photo diodes to detect the alternate flashes, one with a filter (I suspect IR may do), then the code to determine which one is flashing at a given moment and pulse the single LED at the necessary intensity (presumably, pulse width) to provide the counterfeit data to the sensor.

zjjackson101: From what I have read about pulse oximetry, I am having a hard time understanding if i should make a circuit that varies the pulse frequency of the RED and Near IR LED's. Or if I should make a circuit that varies the intensity of the LED's

Your first task is to set up a system to interpret the pattern of excitation of the two LEDs in the sensor.

Since the sensor almost certainly aggregates the input of its own photo-transistor, pulse length will be interpreted as intensity. And since it measures light through fairly dense tissue, you will not need much intensity either.

Paul__B: So you want to create a "dummy finger" eh?

Well (given the detector operates as I say, two LEDs and one photoidiode or photo-transistor), you need a pair of photo diodes to detect the alternate flashes, one with a filter (I suspect IR may do), then the code to determine which one is flashing at a given moment and pulse the single LED at the necessary intensity (presumably, pulse width) to provide the counterfeit data to the sensor. Your first task is to set up a system to interpret the pattern of excitation of the two LEDs in the sensor.

Since the sensor almost certainly aggregates the input of its own photo-transistor, pulse length will be interpreted as intensity. And since it measures light through fairly dense tissue, you will not need much intensity either.

Very Helpful Thank you I will update with my findings.

Any other ideas or suggestions are always welcome!

SantaMedical SM-150 fingertip Pulse Oximeter is a simple device which allows you to monitor pulse rate and SpO2.

johnharada02: SantaMedical SM-150 fingertip Pulse Oximeter is a simple device which allows you to monitor pulse rate and SpO2.

I think the OP's question is more about creating a dummy finger for a device to read, than about the reading.

It would be interesting to see how he progressed in the intervening months, though.

johnharada02: SantaMedical SM-150 fingertip Pulse Oximeter is a simple device which allows you to monitor pulse rate and SpO2.

Yes.

There's a ton of them on eBay, under $12 US.

This is an interesting topic to me, but I have a caveat. I want to know how to monitor your pulse oximetry (pO2) while you are ASLEEP.

Can an inexpensive pulse-oximeter (from eBay) be attached to an Arduino, that can then send data to a computer at selected time intervals, while you are asleep?

This could be very useful for rapidly determining if someone has Obstructive Sleep Apnea (OSA), which causes patients to become hypoxic while asleep.

Thanks for any input.

arbeers: Can an inexpensive pulse-oximeter (from eBay) be attached to an Arduino, that can then send data to a computer at selected time intervals, while you are asleep?

Not really, as they only have an output to the LEDs, which would be very difficult to interface to the Arduino.

What about just using the meter to detect whether you are having a hypoxia event or not? Could the arduino just detect an alert such as a buzz or an LED light?

Thank you for the input and guidance.

There are data logging oximeters. Some may even have serial data output. They are more expensive than the simple finger ones but not outrageously so. You don't even have to lie and say your last name is 'Hospital'.

Getting the measuring clip to stay stable on a sleeping subject is much more difficult.

MorganS: Getting the measuring clip to stay stable on a sleeping subject is much more difficult.

Adhesive tape.

Works real well in theatre! Sticks everything in place!