Minimal oximetery hardware

[zorglub]

I'd like to record multiple sessions of SPO2, perfusion, and pulse rate. Most of the hardware I've looked at is more complicated than I'd like, captures only a single session, and seems to require Windows software. Those nuisances aren't buying me much: I don't need a real time display. I'd therefore like to connect directly to sensor/probe.

Can anyone describe what's involved in talking directly to a sensor, such as the USB probe for the CMS-50F (new style) or CMS-50I at http://www.pulseoxstore.com/Other-Items ... anchor_181 ? Can this (or some similar sensor) be used by DIY folks, or do they all require proprietary drivers that implement obscure protocols?

I interpret the SleepyHead manual's discussion of "live view" as "connected to an oximeter" rather can "connected directly to a sensor". Am I mistaken about that?

[HoseCrusher]

I believe the oximeter reports values by comparing the absorption of two frequencies of light. If you hook directly to the sensors you simply need a comparison look up table for the standard response from the lights used to come up with SpO2 values.

Your best bet may be to look for a fingertip sensor and simply try to communicate with that.

[JDS74]

Oximeters consist of two parts, one the sensor assembly and two, the computer part.
The sensor assembly has a LED emitter and a photo sensor as a pair. Sometimes the photo sensor sees two different light frequencies to help distinguish the correct intensity variation that is used to compute the SPO2 level and also to calculate the pulse rate.

The computer part sees a varying resistance in the photo sensor or varying voltage depending on the type of sensor employed. The firmware then performs a calculation, generally once per second in less expensive models, to convert these levels into a SPO2 measurement. The once per second is the output measurement timing not necessarily the input sampling rate. Using the periodic measurement variation, the firmware also calculates the pulse rate. Together these are combined into pulse-ox pair and displayed and / or recorded.

So to do it yourself, you will need a carefully regulated power supply to supply power to the sensor LED, an algorithm that will allow you to calculate what the sensor resistance / or voltage variation means in term of SPO2 levels, and a separate algorithm that tracks the variation and converts it into a pulse rate.

Then, there is the computing platform where all that runs..

I don't know of a consumer priced device that reports on perfusion.

Personally, I was happy to just pay the $100 and get one already done.

But that is just me.

[zorglub]

I'm not asking how to make a sensor. I don't want to make a sensor either.

I want to use my computer as "the computer part". In other words I want to write my own software that talks to a readily available sensor, such as one of these USB based sensors:
* http://www.pulseoxstore.com/Other-Items ... anchor_181
* http://www.pulseoxstore.com/Other-Items ... anchor_308
* http://contecmedical.en.ec21.com/PC_Bas ... 14790.html

Do these things need some specialized Windows-only drivers, or can I get information from them using some reasonably open API not running on Windows?

[zorglub]

This probe is lower level than I'd like (I'd still need to build some hardware if I use this):
http://dx.com/p/adult-soft-silicone-enc ... vbPdv2OTWw

At least the connector is reasonably standard, even if the pinout is not:
http://pinouts.ru/PortableDevices/nellc ... nout.shtml

The three USB based sensors I already listed are more appealing.

[chunkyfrog]

Couldn't the sensor be linked to an compact Arduino with Bluetooth, eliminating cabling to the computer?

[Lazer1234]

Could this be something?

http://www.cooking-hacks.com/blog/e-hea ... d-in-wired

[zorglub]

Couldn't the sensor be linked to an compact Arduino with Bluetooth, eliminating cabling to the computer?

That would eliminate part of the cabling. A cable between the Arduino and the sensor would still be needed, unless the Arduino with Bluetooth shield is attached directly to the probe. That might not be as clunky as it sounds, with a small Arduino work-alike such as the DigiSpark or the Beetle, and a bluetooth board like http://dx.com/p/wireless-bluetooth-rs23 ... ve3WP2OTWw

Personally, I don't mind the cable. I think it's probably a small inconvenience compared to the mask, the chinstrap, and the guard to prevent tooth grinding.

[zorglub]

Could this be something?

http://www.cooking-hacks.com/blog/e-hea ... d-in-wired

Yes. It's got a pretty good fraction of the sleep study channels, and the complete thing is 450 euros. The cables appear to be too short (especially the oximeter cable). Personally I want the oximeter because that's supposed to be the bottom line of therapy, and I'd like to know how effective it is (averaged over a month or so). I'm not trying to do a DIY sleep study. I'm therefore still more interested in the USB sensor-only.