logo elektroda
logo elektroda
X
logo elektroda

Pulse oximeter: measurement of pulse and oxygen saturation of arterial blood

mb1988 


Hello all.
I would like to present devices that I and two friends made for the subject of Microprocessor Technology at AGH. At the same time, this and my colleagues' applications will be on display at the Krakow Science Festival on May 14 from 11:30 to 16:00, in the main hall of the A0 building at al. Mickiewicz, so I cordially invite you ;-) .

:arrow: Principle of operation
The principle of measuring saturation is based on the fact that oxygenated blood (specifically hemoglobin) absorbs different wavelengths differently. In short, we scan the finger with 660nm and 940nm light, and then we calculate the parameter of interest based on the ratio of the amplitudes of the received signals. The heart rate is calculated based on the time interval between the maxima on the plethysmographic curve.

:arrow: Main board
The system is driven by a 68HC908AB32 microcontroller. Waveforms from the 4-output PWM through a small logical function control the S&H systems and the H bridge switching diodes alternately. The lower transistors in the bridge together with D / A converters are regulated current sources determining the brightness of light pulses. The signal from the photodetector is demultiplexed (red and infrared pulses go to different processing paths), demodulated and filtered. Next, the signal goes to an amplifier with programmable gain to finally be sampled by the transducer contained in the microcontroller structure. The data is presented on a graphic display with 192x64 pixel organization. The keyboard was made on a separate board with illuminated buttons with a label attached to it. These buttons are used to set alarms (pulse decrease / increase, saturation decrease). The whole is powered from a 5V power supply. As you can see in the photo, minor corrections in the schematic diagram were made after soldering the whole :P .


:arrow: Finger sensor
The finger probe was made of - what can I say - oak style from a shovel :P . The wooden piece was cut in half, then hollowed out on a bench drill. The hinge visible in the photo is a formed piece of a large brass wing hinge. The spring comes from "jaws with teeth" hair clips :P . The cable is a memory of my Logitech UltraX keyboard. In the upper part of the sensor there are 4 diodes - 2 red and 2 infrared ones connected in parallel. In the bottom there is a PIN photodiode. In addition, a piece of sponge has been installed in the sensor, which makes the sensor hold securely on both thin and thick fingers.



:arrow: Housing
Due to the fact that the device was intended for a public show, we decided that it should look inviting and distinguish itself somewhat from the usual cold medical equipment. It was made of an aluminum can. The window was cut out with a dremel mounted in a rack. The whole is painted black with a semi-spray. Finally, the graphics were applied with the same hand that adorned the HDDClock standing on my desk ;-) .


Below are some additional photos and a video of the system operation (the alarm was set at 75 beats per minute). For now, I am only providing a schematic diagram (almost final), because the rest of the documentation needs to be worked out before publication. Everyone interesting I invite you once again to the presentation on May 14, and now for discussion.
Regards :) .






About Author
mb1988
mb1988 wrote 93 posts with rating 258 . Been with us since 2006 year.

Comments

parafka 02 May 2010 07:00

Very interesting project. However, I'm afraid that the lack of finger message may slightly distort the pulse rate. :sm31: Can you describe in brief why you need to use two wavelengths of light?... [Read more]

Kuniarz 02 May 2010 08:40

An interesting and well-made project, while on the message "NO FINGER" and the icon of a hand without one finger I would think about it :D Maybe write something like "put on a sensor" etc. [Read more]

piotrva 02 May 2010 08:59

Very interesting design, aesthetically made. Could you just include photos of the "mechanism" of the sensor (optoelectronic components)? As for the message about the lack of a finger, I have no objections... [Read more]

mb1988 02 May 2010 09:04

The message about the lack of a finger is, say, a student, playful approach to the topic - as I said, the equipment will never go to the hospital :P . The measurement is performed for two wavelengths,... [Read more]

elektronik999 02 May 2010 09:09

It is more or less that oxidized blood absorbs one color and blood that has already given oxygen to another. [Read more]

mb1988 02 May 2010 09:17

Exactly. And based on the analysis of signal amplitudes, we can determine how much percentage of iris blood is oxygenated and how much is not. [Read more]

tmf 02 May 2010 09:26

Great design. I have a technical question - why did you use two separate processing paths? It couldn't be done in one? And the second nomenclature note - hemoglobin combined with oxygen is oxygenated... [Read more]

M. S. 02 May 2010 09:32

How did you scale the oxygen content in your blood? Is this measurement independent of finger "parameters" - skin thickness, cleanliness etc.? [Read more]

IoNcs 02 May 2010 09:43

Were the measurements compared with other devices of this type? [Read more]

Anonymous 02 May 2010 09:47

Have you modeled on factory devices? When and where will you be able to buy your pulse oximeter ??? [Read more]

mb1988 02 May 2010 09:52

If I would like to use one processing path, demultiplexing and filtration would have to be done after sampling the signal. The applied process is not a signal processor, so applying filters other than... [Read more]

adamus202 02 May 2010 11:28

And I have a question what is this microcontroller and in what language was the program written for it? [Read more]

Tomekddd 02 May 2010 11:38

Very interesting device ;) I always wondered how it works. I did a "non-invasive blood flow meter" at the ICU and I also wanted to add something to it, but I didn't know how to build it. Now I know... [Read more]

BartekLecki 02 May 2010 11:54

This is Motorola, currently Freescale http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=HC08AB&fsrch=1 in what language (C or assembler) must the author already speak ;)... [Read more]

mb1988 02 May 2010 11:55

Procek is an 8-bit Freescale CISC microcontroller. The program was entirely written in C. The graphics may look illuminated in this light, although this is not the case. It was applied to an already... [Read more]

naelektryzowany 03 May 2010 17:49

Hello, what method were used for the tiles because the effect is phenomenal - no digestion etc. [Read more]

mb1988 03 May 2010 19:42

Plates made by photochemical method. I used Bungard laminate, factory packed with photosensitive varnish. The mask was printed with an inkjet printer on a specially adapted foil. I used a facial tanning... [Read more]

Taenia_Saginata 03 May 2010 20:42

Have you checked how many you can get down to? [Read more]

mb1988 03 May 2010 21:03

The device has software limitations and measures heart rate within 35 - 240 beats per minute, and saturation 75% - 100%. However, being healthy, to get a 75% result I would have to hold my breath for about... [Read more]