1

I am not a expert here. But for a software project of mine I was scratching my head for a wearable device which can monitor heart rate continuously (battery efficient). I have one wearable device Sony MIO fuse which measure heart rate using PPG technology which most of the other wearable device use.

I did some research and found some links: https://ieeexplore.ieee.org/document/5504744/ https://www.computer.org/csdl/proceedings/bsn/2010/4065/00/4065a139-abs.html

which is effective and also battery efficient. Also to listen to heart beat the most ancient technique has been through sound only (stethoscope).

Now my question is why wearable device maker are not using this technique. May be using touch microphone we can listen sounds inside body on wrist. Are there any technical complications to it?

Hope my question is not confusing and I am hoping I am posting it to right forum. Please comment if it need improvement. Like I said before this is not a field of my expertise.

Thanks.

Savaratkar
  • 113
  • 3
  • 1
    excluding external noise may be a big factor... – Solar Mike Jul 25 '18 at 06:37
  • I'd recommend doing a literature search on the topic. Chances are, some folks have investigated the idea and published their difficulties. Such research usually winds up in 4th, 5th, or 6th tier publications -- given that it is "negative research." The top periodicals grab up the headline stuff (successes.) So you will need to use a good search service, I suspect. – jonk Jul 25 '18 at 06:50
  • 2
    Three words: Noise, noise, noise. Have you tried to use microphones to record quiet stuff? Try it. It's ear opening. Our brains are brilliant at filtering it out noise. Much better than we are aware, usually. If you try to record your own heartbeat, amplify it, listen to it with headphones, and look at the waveform, it will all become clear to you. Bonus experiment: Also try doing it while not sitting perfectly still. – Dampmaskin Jul 25 '18 at 07:03
  • For noise reducing I was thinking the method which they use in smartphones. They have an extra microphone at the top of the phone, by software they cancel the sound coming in from this microphone(m1) by the sound coming in from the microphone (m2) in which we speak. m1 microphone in which we speak (+noise), m2 microphone which is there to record noise. – Savaratkar Jul 25 '18 at 08:00
  • If something hasn’t been done it’s either because nobody’s thought of it; the technology’s not developed yet; or it’s been tried and it’s no good. This falls into the third category. – Chu Jul 25 '18 at 10:11

2 Answers2

4

I have worked on this as a hobby project.

It can be done, but getting it to work reliably is a problem.

A microphone and some software can count your pulse. The hard part is doing it reliably in anything but quiet, ideal conditions.

I tried approximately 37 bazzillion ways to tell heart beats from background noise.

Some work well, but depend on a calibration step to get the correct levels - and calibration has to be repeated every time you put the thing on. So if it moves while you are jogging, you get to recalibrate. No fun.

Some methods are insensitive to position and sound level, but are prone to "hearing" heart beats where there are none - like when the microphone is not actually in contact with your body or the microphone is disconnected.

Some other methods work well, but are impractical in an embedded device - too much processing power required.

Most methods have some combination of those (and other) difficulties.

All depend on the heart beat being louder than (or at least distinct from) the background noise - and that's where it all starts.

You need a way to attach the microphone to the body such that it picks up the heart beat clearly, but doesn't pick up breathing noises or motion noises (shirt rubbing on the housing and stuff like that.) It also needs to ignore environmental noises (like the sound of a car driving by as you jog down the road.) And, it needs to ignore the thud of your feet as you jog.

It is complicated, and a working compromise is difficult to achieve and maintain.

Hard enough that I gave up.

JRE
  • 67,678
  • 8
  • 104
  • 179
0

Ok, I know this should just be a comment, but I don't have the rep, and this is too good to pass up.

A few years ago I messed around a bit with taping a piezo transducer disc to my chest, near where I presume my heart ought to be, and observing (with an analog oscilloscope) the signal from the piezo.

Ambient "background" noise was attenuated substantially versus what a microphone would likely capture. I found that if (and only if) I was completely still and literally holding my breath, my heartbeat's signal would be the dominant event on the scope display.