Blood Testing Equipment

I hate to be a misery...but why?

I believe this sort of thing can be bought very cheaply and it will be time consuming and difficult to do it yourself...

Or is this just for the fun of it?

Just programming a micro to talk to and LCD module is a right pain even when you do it for a living!

Yes, you are right, it is much work but this is my first project and it will serve me as a diploma thesis.

I'd start with the mirocontroller, display and buttons,,, If you can get that working the rest should be ok I'd guess...but what is your back ground?

You may end up needing to impement an AtoD converter with the micro or building one in...unless you are an electronics designer it's a huge learning curve.

Where are you?

I study Electrical Engineering and Computer Science in the 4-th year out of 5. I have more than basics in electrical scheme design, so that would not be a problem, but the part that concerns me is the sensor because I don't know how to choose it. I will use the internal 10bit A/D converter of the microcontroller.

i remember that in medical devices there were some light sources and then after the light passes through the sample probe it's wavelength changes, there is a sensor that reads that wavelength. I would choose something in that area, but there might be that I could find a more practical solution which I am hoping to find.

Sorry I'm no help on the sensors...

Good luck & have fun.

Thank you for your answers, i hope someone will help me on the sensor part and after that i will address the chemical forum questions about how to prepare or procure the reagents.

As a person with Diabetes we are given FOC the blood testing equipment; but you can buy one for £8.00 and the test strips are FOC if prescribed by your Doctor. It is similar for my INR aka how thick is my blood that Monitor cost me £400 [because I'm a fool for working, if on benefits it would be free] but the testing strips are FOC if prescribed by your Doctor and its similar for cholesterol testing you can buy these on the high street the last time I brought a friend one from boots [chemist] it was £8.50.

I am interesed to build a device that measures: glucose, cholesterol, triglycerides, and some other basic blood tests, maybe even calcium concentration. The cost won't be much over 50$ and if I do it right it might be even cheaper.

Hi Chris, have a look at, for example, the Ocean Optics website to see what's already been done regarding optical blood sensors and don't be too worried about reinventing the wheel as it is the hands-on practical experience, with hopefully, the satisfaction of getting the thing to work reproducibly (not just once!) that counts. Good luck.

thank you, I had a look at that website but was horrified by the prices. I want to build something in the range of 50$, where only the lamp and a spectrometer were ~3000$. I found some wavelength sensors and now I am searching what a light source i can use, that could be also reliable, not fading away.

I work in the business of medical devices that do pretty much what you want to. I think the spectrometer is the most important thing - and not cheap since it is the heart of the system. It would be nice to do it simpler, easier, cheaper but I don't think it can be done. The white LED (quality has improved greatly!!) and filtered sensors may give a start, but it would be very crude. There are a lot of fine details in analyzing chemisties like this. ss

I think that for a sensor I will be using something like this:

http://www.advancedphotonix.com/ap_products/pdfs/SD100-11-31-221.pdf but for the source light, how can i be sure that a LED contains the whole spectrum, or at least the range 350-800 nm? This machine will be something to grow into, so after this version i will try updating it gradually. I thought on finding LED's with fixed characteristic wavelength, one LED for each test but I am not sure that I can find them so specific.

Look in Farnell or your local catalogue or manuf's data sheets...most leds specify a wavelength...

or they did when I last looked..

(I reserve the right to be wrong)

Once you select your sensors, maybe you can use a National Instruments device and a LabVIEW program to run your test using your desktop/laptop.

This is entirely off the top of my head... but I'd think a white led might be a good light source, if the system self-tested for white balance at each startup. Three phototransistors with r,g, and b filters to measure color? CCD and some of the electronics from a small camera?

the phototransistors I found on digikey.com were covering only the 800-940 spectrum.so, i think that the wavelength sensor would still be the best choice.