Electret mic and phone for heart and lung sound for telemedicine

[Buk]

Now I show the perfect low distortion sinewaves at the output of my circuit

3 problems with that.

1. Heartbeats are not perfect sinewaves; and certainly not 1kHz sinwaves. They are a mismash of frequencies ranging between ~10Hz and ~1000Hz; all mixed together.
2. Your graphs -- I assume -- are the output waveforms; showing roughly 7V peak to peak; but the input stages of the phone are expecting the input from an unamplified electret mic. That's why everything is being clipped. Your circuit is designed to drive the 8 or 16 ohm input of a pair of headphones or speakers, not the input stages of smartphones.
3. Smart phones need to see an impedance of 800+ ohms(*)to recognise that an external mic is plugged in.

* Source of information
Further evidence:

 

[Buk]

By way of explanation of the sort of thing I think is required; a simple, ~2x amplification using a mosfet. Note I'm not recommending this as an actual solution, but it at least demonstrates the veracity of the simulator.

Note: how the input (lower trace) is somewhat less than 200mV peak to peak.
Note the output (upper trace) into 2.2k Ohms is ~350mV peak to peak.
Note how closely the amplified output waveform (upper green) tracks the input waveform (lower green).
Note how closeley the spectrum of the output matches that of the input.

It would need a low-pass filter; and probably a switchable or variable pass frequency depending upon heart sounds or breath sounds.

Feel free to tear the circuit apart; but please at least offer explanations and corrections.

 

[audioguru]

The 2nd-order Butterworth lowpass filter (two 47nF capacitors) in my original heartbeat stethoscope had a -3dB cutoff at 100Hz. I modified it (two 1.5nF capacitors) to hear breathing sounds in this thread with a cutoff frequency of 2.5kHz. A switch can easily be added.
My original circuit used an LM386 power amplifier IC with a gain of 20 times and a volume control to drive low impedance headphones or maybe a speaker. It clearly produced the heartbeats.

The circuit in this thread used a 1Meg pot for an adjustable total gain from 1.6 times to 35.6 times.

It would be difficult to prevent hearing heartbeat harmonics when listening to breathing sounds if the circuit has an added highpass filter making a bandpass filter.

 

[Buk]

It would be difficult to prevent hearing heartbeat harmonics when listening to breathing sounds if the circuit has an added highpass filter making a bandpass filter.

Sorry, I wasn't suggesting a high-pass or band-pass. Only a variable/switchable low-pass.

The idea -- based more on limited observation of available recordings, than any actual knowledge -- is that lowering the cut-off frequency when listening to heartbeats might prevent them being swamped by the generally higher frequency and naturally more intense breath sounds.

When listening to breath sounds the cut off frequency would raised, and a combination of: the naturally greater intensity of the higher frequencies, the inherent tendancy of phone amplifiers to favour higher frequencies, and the physical repositioning of the stethoscope head; would serve emphasis the breath sounds.

Whilst this could be done in hardware, I think it would increase both costs and complexity and probably risk introducing unwanted harmonic distortions. The purpose might be more easily served and better controlled by implementing the filter in software within an app written for the purpose. There are many such apps already available, Many for free. The wikipedia article says:

Because the sounds are transmitted electronically, an electronic stethoscope can be a wireless device, can be a recording device, and can provide noise reduction, signal enhancement, and both visual and audio output. Around 2001, Stethographics introduced PC-based software which enabled a phonocardiograph, graphic representation of cardiologic and pulmonologic sounds to be generated, and interpreted according to related algorithms. All of these features are helpful for purposes of telemedicine (remote diagnosis) and teaching. Electronic stethoscopes are also used with computer-aided auscultation programs to analyze the recorded heart sounds pathological or innocent heart murmurs.

That said, this quote from the wikipedia article on stethoscopes -- especially the enboldened part -- is something for doldett to consider.

An electronic stethoscope (or stethophone) overcomes the low sound levels by electronically amplifying body sounds. However, amplification of stethoscope contact artifacts, and component cutoffs (frequency response thresholds of electronic stethoscope microphones, pre-amps, amps, and speakers) limit electronically amplified stethoscopes' overall utility by amplifying mid-range sounds, while simultaneously attenuating high- and low- frequency range sounds. Currently, a number of companies offer electronic stethoscopes. Electronic stethoscopes require conversion of acoustic sound waves to electrical signals which can then be amplified and processed for optimal listening. Unlike acoustic stethoscopes, which are all based on the same physics, transducers in electronic stethoscopes vary widely. The simplest and least effective method of sound detection is achieved by placing a microphone in the chestpiece. This method suffers from ambient noise interference and has fallen out of favor. Another method, used in Welch-Allyn's Meditron stethoscope, comprises placement of a piezoelectric crystal at the head of a metal shaft, the bottom of the shaft making contact with a diaphragm. 3M also uses a piezo-electric crystal placed within foam behind a thick rubber-like diaphragm. The Thinklabs' Rhythm 32 uses an electromagnetic diaphragm with a conductive inner surface to form a capacitive sensor. This diaphragm responds to sound waves, with changes in an electric field replacing changes in air pressure.

 

[rjenkinsgb]

Another method, used in Welch-Allyn's Meditron stethoscope, comprises placement of a piezoelectric crystal at the head of a metal shaft, the bottom of the shaft making contact with a diaphragm.

Interesting.. That sounds vaguely similar to the construction of a "drum trigger" using a piezo disc and foam cone as a contact point.



Those are designed to stand high impacts and long movements of the surface being sensed, so have a very tall foam block to absorb the impact.

Piezo discs are available in various diameters and very cheap - around 20p - 40p or so each when bought in 10s.

10Pcs 20mm Piezo Elements Buzzer Sounder Sensor Trigger Drum Disc+ wire copp GF | eBay

Find many great new & used options and get the best deals for 10Pcs 20mm Piezo Elements Buzzer Sounder Sensor Trigger Drum Disc+ wire copp GF at the best online prices at eBay! Free delivery for many products!

www.ebay.co.uk

Pardon our interruption...

A piezo disc mounted in a shallow container such as a bottle cap, with a small self-adhesive foam pad eg. 3mm thick in the centre, just to give slight conformance to the skin, could give a similar effect as the piezo stethoscope pickup in the link?

You would need a high impedance preamp or buffer, but the output level should be quite high - it's in the volts to tens of volts range when piezo discs are used in drums!

 

[audioguru]

My electronic stethoscope circuit with an electret mic and LM386 power amplifier produced very clear heartbeats in my headphones.
I did not try hearing breathing sounds with a higher frequency cutoff of the lowpass filter because I am a healthy non-smoker with no breathing sounds.
I never use a piezo transducer as a microphone because its frequency response is horrible.

 

[Buk]

with a small self-adhesive foam pad eg. 3mm thick in the centre, just to give slight conformance to the skin, could give a similar effect as the piezo stethoscope pickup in the link?

The choice of foam would be critical. Both open-celled and closed cell foams are used as acoustic foam; but for controlling different frequencies.

There is also the possibility of using a hard contact with the sternum and using it as a sounding board, though that might introduce resonances that would muddy the signal. It would also probably be quite uncomfortabel to have some hard pressed against the bone.

There are many other possibilities. It really depends upon how urgently doldett needs to finish; and how much time and money he is prepared to invest. The classic trade off of some very good or something good enough. Those companies mentioned probably invested considerable R&D into their solutions.

One of the most interest notions I've had is the idea of using two omni-directional electrets back-to-back and differencing their signals to make a virtual cartoid mic, thus eliminate background noise in a classic noise-cacelling fashion. Taken a little further, mounting the forward facing electret in an appropriate foam to isolate it from the housing and mounting the backward facing electret direct to the housing ought to allow a large portion of handling noise to be eliminated.

Whatever way he goes, I think his priority should be to get as much signal as he can into the phone with the minimum of hardware manipulation beyond elimination background & handling noise. Once the signal is recorded, any amount of filtering and selective amplification can be done in software, with the bonus that once recorded, you can try several different filtering and enhancing strategies because signal is recorded. And of course, add improvements after teh hardware is in the field by downloading a new version of the app.

 

[doldett]

Electronic stethoscopes require conversion of acoustic sound waves to electrical signals which can then be amplified and processed for optimal listening. Unlike acoustic stethoscopes, which are all based on the same physics, transducers in electronic stethoscopes vary widely. The simplest and least effective method of sound detection is achieved by placing a microphone in the chestpiece. This method suffers from ambient noise interference and has fallen out of favor. Another method, used in Welch-Allyn's Meditron stethoscope, comprises placement of a piezoelectric crystal at the head of a metal shaft, the bottom of the shaft making contact with a diaphragm. 3M also uses a piezo-electric crystal placed within foam behind a thick rubber-like diaphragm. The Thinklabs' Rhythm 32 uses an electromagnetic diaphragm with a conductive inner surface to form a capacitive sensor. This diaphragm responds to sound waves, with changes in an electric field replacing changes in air pressure

Very interesting indeed. I thought piezo-disc generating electrical signal in response to its deformation would not be applicable with such low acoustic sound wave. It, however, definitely needs to be looked at further if piezo-disc is used in commercial electronic stethoscope.

There are many other possibilities. It really depends upon how urgently doldett needs to finish; and how much time and money he is prepared to invest. The classic trade off of some very good or something good enough. Those companies mentioned probably invested considerable R&D into their solutions.

Buk, my friend, who is a doctor, thinks we need to finish first prototype within the next few weeks just to show that it is capable of listening breathing sound through video call application. If our first prototype is successful, we might seek for financial support from one of the government agencies. It is gonna be hard to get financial support, but we are gonna try anyway. I also think I am gonna pursue this endeavor even when I do not get the financial support ^_^.

 

[Buk]

My electronic stethoscope circuit with an electret mic and LM386 power amplifier produced very clear heartbeats in my headphones.
I did not try hearing breathing sounds with a higher frequency cutoff of the lowpass filter because I am a healthy non-smoker with no breathing sounds.
I never use a piezo transducer as a microphone because its frequency response is horrible.

An article picked from a search engine because it was near the top -- so no guarantees about its efficacy -- but an interesting read.

"Piezo mics aren’t usually considered hi-fi, so distortion isn’t usually a problem of the amplifier – at high levels the mic will not be particularly linear. However, they are remarkably low noise. Although they are high impedance, they don’t need an outrageously high load impedance – not like an electret mic, where you need gigaohms to avoid losing bass. The calculation in the introduction showed a load of 330k was enough, so 1 meg is plenty for a contact mic. "

 

[rjenkinsgb]

I could not resist testing that piezo idea..



It's a really basic, nasty setup - a plastic bottle cap with an offcut of thick wire fitted around the inside to support the edge of the piezo disc and a couple of bits of self-adhesive polyurethane foam panel gasket as a centre actuator, then tape over the wires at the back to stop it falling apart!

It works. Fed in to an audio interface set in "instrument" mode (eg. for a high impedance guitar pickup), it strongly picks up my heart when held against my chest. The output is almost entirely low frequency; not very audible but it makes my sub overtravel and thump if I turn the volume up much!.

The only processing was a 1KHz low pass before exporting it as a .wav, which actually did not make much difference.

I also tried a 50Hz low pass (-60db) out of curiosity & that also did not have much audible or visual effect on the waveform; I did not save that.

I suspect a much heavier casing, to add inertia and better restrain the disc edge, would improve output and reduce handling noise. I was holding it against my chest with one hand and operating the mouse with the other, so it's picking up movements from the hand holding it, the hanging cable that connects it & possibly a delayed pulse from my fingertips. Not an ideal test setup - but it's a proof of concept.

The .wav sample is here linked below; be careful with amp volume, use headphones, starting at low volume if you cannot hear anything . It may be inaudible on normal speakers but will send a subwoofer crazy..

**broken link removed**


Edit - More thoughts.
A small piece of viscous "memory foam" may work well for the centre actuator? That should conform to the skin easily, but pass relatively rapid movements rather better than simple elastic foam.

Another idea, considering the output is such low frequency - use it to frequency modulate an audio tone, something like with slow-scan TV, then demodulate for plotting or oscilloscope style display at the other end of the connection.

 

[audioguru]

Buk, the article you found is completely wrong about the electret mic and its low frequency response. It describes a condenser mic which is an electret mic minus the Jfet inside the electret mic. The Jfet in the electret mic already has the 1G ohms resistor biasing it. The article uses an old NE5534 opamp that has a minimum input resistance of only 30k ohms but at least 1M is needed. The OPA134 opamp I recommended has Jfet extremely high resistance inputs.

RJenkins, I could barely hear your piezo heartbeat but with headphones I heard it. It works!
But yeah the background and rubbing sounds were bad. Mounted with a rubbery adhesive in a rigid cap it will be fine to hear heartbeats but not problems or lung sounds.
The frequency response of a piezo transducer is so bad that heartbeat problem sounds or lung sounds will be extremely distorted.
A piezo produces a very high voltage when knocked (BBQ lighter) that can zap an amplifier. A Piezo drum pickup uses diodes to clamp the high voltages.

 

[Buk]

The .wav sample is here linked below

There is an aweful lot about that to like. It has a very strong signal right across the range that is needed.

This is it next to the the very clean 16 y/o heart from wikipedia:

If anything I'd say that you are overdriving the amp, perhaps try a lower voltage.

Also, try to mechanically isolate the pickup from handling. First pass simple suggestion would be something like this:
The pink is a rubber (glove?) membrane sandwiched (and glued) between two close-fitting peices of plastic (green) (bottle top and neck of the bottle it came from); the grey is a billet of something heavy -- steel/brass/UHMW PE -- glued to the membrane; the gold/silver/cream is the peizo glued to the billet; and the fuzzy grey is something comfortable to contact the skin. Could be foam, but I think silicone rubber would be better for sound transmission.

 

[Buk]

Another idea, considering the output is such low frequency - use it to frequency modulate an audio tone, something like with slow-scan TV, then demodulate for plotting or oscilloscope style display at the other end of the connection.

I just caught your update. Could you expand on it a little?

I liberated a 27mm peizo unit from a dead dect phone last night and I'm working on finding a suitable rubber for my membrane. I've gathered together materials for the other parts of my steth head design, but they'll need some fettling.

 

[rjenkinsgb]

Slow scan uses audio FM; the audio tone varies from around 1200Hz to 2300Hz representing minimum to maximum signal voltage or level.

A similar system could be used to send low frequency waveforms that cannot pass through a normal telephone audio link; eg. around 1800Hz for zero volts and 1200 - 2400 for most negative and most positive.

It could be used for accurately sending the very low frequency heart waveforms or sounds, up to a few hundred Hz.

Digital versions of audio FM (Kansas City / CUTS) were used to store computer programs on audio tape, using 1200/2400 Hz tones at anything from 300 to 1200 Baud.

More info:

Slow-scan television - Wikipedia

en.wikipedia.org

Kansas City standard - Wikipedia

en.wikipedia.org

 

[audioguru]

Doesn't this project have a very low cost budget so that sick poor people in a poor country can buy an inexpensive product that can send their heartbeat and breathing sounds from their home to a doctor's office? An ordinary phone line cannot pass the extremely low frequencies of heartbeats so the modulation idea will work very well and can use AM or FM modulation. Unfortunately this product will probably be too expensive.

During a heart attack I had an angioplasty operation with angiograph video and after surviving I had many stethoscope and ECG tests in a doctor's office. I carried and slept with a very expensive portable ECG device and then a portable blood pressure device. My government paid a fortune for it all. Now I am better than new.

 

[Buk]

An ordinary phone line cannot pass the extremely low frequencies of heartbeats

But they can upload a .wav file containing the extremely low frequencies of heartbeats.

 

[audioguru]

People who cannot pay to go to a hospital or doctors office have a smart phone in a poor country that has no cell phone or internet networks?
EDIT: Sorry, I understood it wrong. Maybe the people and country are not poor, they have a Covid-19 infection and are at home in quarantine.

Here with the government paying for health care, the infected people are taken to a hospital in an ambulance.

 

[Buk]

A similar system could be used to send low frequency waveforms that cannot pass through a normal telephone audio link;

So, use the voltage output from the piezo to adjust the signal frequency of an FM generator?

 

[Buk]

People who cannot pay to go to a hospital or doctors office have a smart phone in a poor country that has no cell phone or internet networks?
EDIT: Sorry, I understood it wrong. Maybe the people and country are not poor, they have a Covid-19 infection and are at home in quarantine.

Here with the government paying for health care, the infected people are taken to a hospital in an ambulance.

I'm not sure that I understand the intent of your question?

My understanding of doldett's purpose was to allow people in Thailand to have a consultation with an appropriate medical professional without needing to leave their home during lockdown due to the current (and almost enevitable future) pandemic(s).

My point was simply that any such tele-consultation doesn't need to be conducted live. A transmitted file could be reviewed by the consultant and only those with cause for concern be followed up with live communications.

 

[doldett]

My understanding of doldett's purpose was to allow people in Thailand to have a consultation with an appropriate medical professional without needing to leave their home during lockdown due to the current (and almost enevitable future) pandemic(s).

My point was simply that any such tele-consultation doesn't need to be conducted live. A transmitted file could be reviewed by the consultant and only those with cause for concern be followed up with live communications.

Our first intention is for doctor to hear live breathing sound from covid patients. I believe that the recording of lung sound should be fine as well.