Looking for a device to trigger software off my pulse

April 26, 2008 8:05 PM

Following up on my previous question, I'm looking for a good way to have someone's heartbeat trigger software on a PC.

Previously, I asked about building a digital stethoscope for use in triggering an interactive sculpture. I've gotten that working pretty well, but it turns out it's pretty sensitive to noise. In a quiet room it works great; our software can pick up every heartbeat and trigger on it very reliably. But we set up in a noisy gallery today and the noise picked up by the stethoscope definitely degrades performance.

So we'd like to try some non-acoustic methods to measure heartbeat. We like the idea of using a pulse oximeter, but none of the ones I saw googling seem to output an actual pulse signal - they just give you heart rate. Ideally we'd like a device that outputs the actual waveform from the sensor or just a digital signal every heartbeat.

I'm open to building something if that's what is necessary so ideas like moonmilk's previous comment or this are useful too.

posted by pombe to technology (6 comments total)

I'd check if the GPS training systems allow for live feed to a computer - with the Garmin models you can send your heart rate data to your computer once your training is completed, but I'm not sure it can be done live.
This might also be of interest to you.
posted by ddaavviidd at 8:18 PM on April 26, 2008 [1 favorite has favorites]

Making a working ECG can be pretty challenging, and based on a student who's trying to do it for his final project, depends a lot on good electrode-to-skin contact (meaning, each user needs a few dabs of conductive goop on their skin). Have you tried to insulate your stethoscope from any noises that aren't coming from your victim's chest? (This is easier if your stethoscope doesn't have to look like the doctor's but could be, say, a tube-in-a-tube to allow for some insulating sheathing.

There's also the solution from Polar, but that doesn't give an ECG waveform either, just a 1ms pulse for every heartbeat.

Another solution that comes to mind would be to cheat -- if you're in the market for a pulse oximeter or the Polar sensor, you could use the rate data from that to set the playback speed for a pre-fab ECG waveform (in software). I suppose this runs the risk of a heart patient using your device and "discovering" that their heart murmur has disappeared, so they stop taking their pills, etc.
posted by range at 9:06 PM on April 26, 2008

A few days ago I was lying in the sun in my hammock with my ipod on my belly, and I saw a spot of light on a nearby wall bouncing in time with my heartbeat. It was the sun, reflecting off the shiny back of the ipod. I can't say, though, that I can see a practical way for you to adapt this technique.
posted by moonmilk at 9:27 PM on April 26, 2008

More helpfully, some googling around found this - a pulse oximeter with an RS232 interface, protocol document, and sample host software.

You could also try looking for devices available from a drugstore, amazon, etc. which feature an LED that blinks in time with the heartbeat. You could cut the LED's leads and send the signal to your computer instead.
posted by moonmilk at 9:35 PM on April 26, 2008

Um... to me it sounds like you need a little signal processing. a low pass filter might really clean it up and it is easy to build. Or you could use some digital stuff.

I've worked with a nice little free piece of software called vvvv, which allows some degree of real-ish time signal processing you could try cleaning up the signal through that.

I'd try some band pass filtering and maybe a noise gate.
posted by magikker at 12:27 AM on April 27, 2008

If you are getting too much noise when it is off someone's chest you could put a contact on the scope so that the mic is only on when it is on someones chest.

You might also want consider other mic elements. Think about the difference between a dynamic mic and a condenser mic. If you don't know the difference find a friend that records stuff and have them show you. But basically condenser mics are used in the studio and are really sensitive. Dynamic mics are used on stage, because they don't pick up as much of the rest of the noise on the stage coming from guitars and such. I think you've got a condenser on your hands when really you'd be better off with a dynamic. Dynamic shouldn't pick up as much back ground noise.

Other option would be to hack this thing, and just use the finger part. Chances are you could un plug that cable, send power to the right parts and get a signal back.

One last random idea would be to use a piezo element like they use to amp an acoustic guitar. I think you can make out of a 5$ piezo buzzer from radio shack. Just need a pre-amp. Instructables has how to built both, at least I think they do. You might be able to use it with the scope... or if they are really sensitive tape it to your wrist or shest or something.
posted by magikker at 12:52 AM on April 27, 2008