Best microphone to detect heartbeat

Thread Starter

icydash

Joined Jan 14, 2009
124
Hi,

I'm trying to build a wearable device that uses a microphone (as opposed to typical optical sensors) to detect my heartbeat. I would appreciate your help in identifying a good microphone for the job. I tried the INMP401 (a MEMS microphone, link), but even pressed directly against my skin over my heart, it was not sensitive enough to pick up my heartbeat. It mostly captured background noise from my TV.

I think the ideal microphone would be small and low power (since it's going into a wearable), directional to limit background noise, and have high enough sensitivity / low enough frequency response to capture the sound of my heartbeat. In actual use, the microphone will not be pressed directly against my skin, but will be positioned against my t-shirt over my heart, so it will probably need to have pretty good sensitivity to hear my heartbeat through my t-shirt.

Thanks in advance.
 

Alec_t

Joined Sep 17, 2013
11,400
In actual use, the microphone will not be pressed directly against my skin, but will be positioned against my t-shirt over my heart
I think you'll find the rustling of clothing will drown out the heartbeat, but I'd like to be proved wrong.
 

ebp

Joined Feb 8, 2018
2,332
It must have good low frequency response. If you look at a typical dual-head stethoscope, the "bell" side will be about 2.5 cm in diameter the diaphragm side will be about 4 cm (infant scopes are smaller). That is a huge area compared with the diaphragm of of tiny MEMS mics. To use a MEMS mic, you would need the equivalent of the stethoscope head. The diaphragm side works reasonably well through thin clothing, the bell almost not at all. Even the diaphragm will get nothing if there is not actual contact - a tiny air gap makes it useless.
 

Thread Starter

icydash

Joined Jan 14, 2009
124
Thanks for the tips so far, everyone!

Do you have any suggestions on microphones that might work? Doesn't need to be a MEMS microphone. I'm interested in whatever solutions you can provide :)

If it's not possible to do this without having the microphone directly against the skin (e.g., over clothing) due to the noise from the rustling of the clothing, that's not as optimal, but I can make that work.
 

iONic

Joined Nov 16, 2007
1,650
Some active filtering with whatever microphone you do choose might be helpful, but I would not be the one to guide you through that process.
The optical method of heartrate/pulserate monitoring really is the better way to go.
 

Thread Starter

icydash

Joined Jan 14, 2009
124
I'm also not opposed to making some kind of membrane like on a stethoscope and attaching a mems microphone to it, if that's what it takes. Looking for creative solutions :)

Unfortunately I can't use the optical method of heartrate detection for this project, due to other constraints.
 

AnalogKid

Joined Aug 1, 2013
8,487
Picking out the heartbeat *acoustically* will be extremely difficult. Since it sounds like direct contact is allowed, consider other technologies. My Polar heart monitor is low cost and transmits the heartbeat by radio to a compatible receiver. Old tech, very reliable.

ak
 

Alec_t

Joined Sep 17, 2013
11,400
What about using a piezoelectric contact mic, instead? Like this:
Worth a try, but (a) at 26mm diameter it's hardly 'small' and (b) its resonant frequency is 4.6kHz, so it's unlikely to be very sensitive at typical heartbeat frequencies.
Have you tried a bog-standard electret mic?
 
Last edited:

Thread Starter

icydash

Joined Jan 14, 2009
124
Worth a try, but (a) at 26mm diameter it's hardly 'small' and (b) its resonant frequency is 4.6kHz, so it's unlikely to be very sensitive at typical heartbeat frequencies.
Have you tried a bog-standard electret mic?
26mm is only about 1 inch right? I think that'll still be small enough for my wearable application. As long as the wearable is like, less than 2 inches in diameter when complete, that's fine. Any ideas where I could find a piezo disc with a much lower resonant frequency? I'm not seeing one on Amazon, but I can't search by resonant frequency so I might have missed something.

I hadn't tried electret mics yet, but I bought these last night: https://www.amazon.com/gp/product/B00OK60U5A/ref=oh_aui_detailpage_o01_s01?ie=UTF8&psc=1 ... Are those what you meant?

I'll let you know how they fare :) Thanks for the suggestion.
 
Last edited:

spinnaker

Joined Oct 29, 2009
7,835
There is a reason electronic stethoscopes are expensive.

https://ekohealth.com/shop/eko-core/?utm_source=Bing&utm_medium=cpc&utm_campaign=Electronic-Stethoscope&msclkid=27cc6c5585ac1230b016307d68853e6e

Find a cheaper one and try to hack it.

Unless you have a lot of experience in design, trying to do it yourself is going to be extremely difficult. The microphone part is the las of your worries. Filtering ambient noise is going to be the biggest part of the problem. Figure out how you are going to do that before searching for a micropjome.
 

Thread Starter

icydash

Joined Jan 14, 2009
124
There is a reason electronic stethoscopes are expensive.

https://ekohealth.com/shop/eko-core/?utm_source=Bing&utm_medium=cpc&utm_campaign=Electronic-Stethoscope&msclkid=27cc6c5585ac1230b016307d68853e6e

Find a cheaper one and try to hack it.

Unless you have a lot of experience in design, trying to do it yourself is going to be extremely difficult. The microphone part is the las of your worries. Filtering ambient noise is going to be the biggest part of the problem. Figure out how you are going to do that before searching for a micropjome.
I guess that's why I want to get the microphone first, so I can start recording data in Audacity on my laptop and trying to figure out exactly what I'm dealing with, noise-wise. My thought is that, until I get a sense of what the microphone's response is like and what it's picking up, it'll be hard to figure out what type of filtering I might need.
 
Using a microphone to detect heartbeat is certainly possible. For example, look at this iPhone app that does just that http://bluespark.co.nz/heartmonitor.php

There is even some early spirometery applications for smart phones https://ubicomplab.cs.washington.edu/pdfs/spirosmart.pdf

The part that makes it ill-advised is when you say
a wearable device that uses a microphone (as opposed to typical optical sensors) to detect my heartbeat.
You now have a lot of other sound to deal with essentially making a problematic signal to noise ratio, if by wearable you mean literally wearable while you go about normal activity. That situation is in contrast to using a stethoscope under controlled conditions, for example.

The wearable approaches that are the most successful use either the optical methods (like the fitness bands) or the ECG (electrodes to pick up the electrical activity of the heart) method (like a Polar chest strap - see this experimenter's kit https://www.adafruit.com/product/1077).

Using a microphone under normal activity conditions is going to be a real challenge when compared to optical or ECG methods, which is why you do not see such applications very often.
 

spinnaker

Joined Oct 29, 2009
7,835
Using a microphone to detect heartbeat is certainly possible. For example, look at this iPhone app that does just that http://bluespark.co.nz/heartmonitor.php

There is even some early spirometery applications for smart phones https://ubicomplab.cs.washington.edu/pdfs/spirosmart.pdf
That is a spirometer. There is nothing acoustic about it. They are simply using the microphone as an easy way to get data from the spirometer to the phone. It avoids a bluetooth connection. There are various devices that use this method,
 
That is a spirometer. There is nothing acoustic about it. They are simply using the microphone as an easy way to get data from the spirometer to the phone. It avoids a bluetooth connection. There are various devices that use this method,
I mentioned the spirometery application explicitly (the second link). The first link in what you quoted from my post has this:


along with the text:

Because Heart Monitor uses sound to detect heart rates, you can use the iPhone 3G directly over your heart or directly on the pulse in your neck or wrist. Heart Monitor works best in quiet areas. You can get excellent results when using the microphone on your headphone cable.


So, no I don't agree with your conclusion that:
There is nothing acoustic about it.
But, maybe you know much more about it than I do, so please do explain.
 

Thread Starter

icydash

Joined Jan 14, 2009
124
Raymond - Interesting. Any idea what type of microphone is in the iPhone headphone cable?

I'm OK with this being a challenging project. If it was easy, it wouldn't be fun and I wouldn't be on here picking the brains of you smart people :)

Alec - Good to know that soundcards have a huge drop off at 20Hz. I wouldn't have ever thought of that. I wonder if buying an external USB soundcard stick (like this: https://www.amazon.com/Sabrent-External-Adapter-Windows-AU-MMSA/dp/B00IRVQ0F8/) would have better response than whatever crappy soundcard is built into my laptop.
 

spinnaker

Joined Oct 29, 2009
7,835
I


So, no I don't agree with your conclusion that:


But, maybe you know much more about it than I do, so please do explain.
In the case of the spirometer, it is simply sending serial data to the phone to be processed by the phone via the microphone connection. All of the measurement is being done by the device you see the person blowing into. The device is not listening to your breath sounds. It is measuring it and turning that measurement into data to be processed by the phone. There are a number of devices that use this trick.

You don't even need a phone to make it work if you were willing to design a MCU project to make it work.
 

Alec_t

Joined Sep 17, 2013
11,400
I suspect the smartphone apps (some of them at least) for heartrate are using the device's accelerometer for input. Have you considered using a MEMS accelerometer?
 
In the case of the spirometer, it is simply sending serial data to the phone to be processed by the phone via the microphone connection. All of the measurement is being done by the device you see the person blowing into. The device is not listening to your breath sounds. It is measuring it and turning that measurement into data to be processed by the phone. There are a number of devices that use this trick.

You don't even need a phone to make it work if you were willing to design a MCU project to make it work.
Both links are applications that depend on acoustic input to the microphone. The first one that I linked to represents detecting a heart rate based on acoustic information - exactly what the OP wants to do. The second application paper that I linked to also uses acoustics from your breath picked up by the microphone. It is why I mentioned it as a second example of a microphone being used to measure another bio-process

When you say, "All of the measurement is being done by the device you see the person blowing into. The device is not listening to your breath sounds. It is measuring it and turning that measurement into data to be processed by the phone. " you are completely wrong and I suspect that you either did not read the linked paper or did not understand it.

They are analyzing the SOUND of your exhaled breath - that is the whole point. I have no position at all about how well or how poorly it works - obviously the authors think that it can work. As far as I am concerned, validation is yet to come and that is why I used the phrase "early spirometery applications".

The basics of their "algorithm" are given in Figure 4 (reproduced below) and Figure 5.

Spiro.jpg

Additionally, Figure 1. shows the use of the app (left: notice NO other devices) and then the conventional spirometer (right) along with the generated curves on the bottom (left and right) for comparison. The mouth piece seen in Figure 3, as stated clearly in the paper, was used during positional standardization and training.
 
Top