DIY 'ECG' Machine: Testing Feasibility of Wave Transference

[DrBanana]

I'm trying to make a DIY 'ECG' machine, except that it'll only record heartbeats on a piece of paper. Basically the piece of paper will be wound round a cylinder like object, which will be being spun slowly by a motor. A pencil at the end of a stick or something will be writing to this piece of paper. A section of this stick will rest on top of a funnel that's covered by a cut out baloon, so if that membrane 'jumps', so will the stick, and the reading will be visible on the piece of paper. The funnel will be connected with a tube to another funnel that's also covered by a cut out baloon, and that will be placed on the chest.

So my question is, is this feasible? The chest doesn't really 'jump' that much so even if I did manage to find the perfect tube, would I be able to make the stick jump to a significant degree? I'm hoping that I can do that if I can manage to find a long enough stick (because the 'jump'=hsin(x) increases as the length of the stick h increases, if I'm not mistaken).

So now I need to find a tube that can best carry waves. I haven't really read in depth about elasticity or waves, but if I had to guess I'd say that materials with a higher Young's modulus tend to carry sound waves better. However here I have another question: are sound waves and waves that make things 'jump' identical? I.e. if I have a medium that carries sound well, does it carry the 'jumping' part well? I've bought a section of a hose pipe but haven't gotten to try it out yet, but I don't have much hopes for this.

 

[berkeman]

It sounds like you'd prefer to keep this all mechanical, and not add any electrical amplification, right? That will be a challenge, since there is so little sound energy from the heartbeat when auscultated (listened to) on the surface of the chest. A stethoscope is the best pickup device that we have for that, and it is all sound-powered, but the energy available at the earpieces is very low (just enough so that you can hear the heartbeats with the earpieces tight in your ear canals).

But you really aren't trying to record any ECG waveforms anyway, since that involves the electrical activity of the heart as sensed through the skin of the chest. Your setup is really only listening to the pulse activity of the heart, not recording any ECG activity.

So my suggestion for a better way to try to do such a recording of the pulse mechanically would be to come up with a probe tip that you can locate over the radial artery in the wrist, and record the small bounces/movement as the pulse makes the radial artery expand with each heartbeat. You can feel your own pulse as shown in the diagram below -- the radial artery is located on the under (anterior*) side of the wrist, between the radius bone and the tendons in the middle of the wrist. Use 2-3 fingertips to press gently into that notch, and find where you can feel your pulse. Then try that on a few of your family members and close friends, to get an idea for the variations in where to find that radial pulse.

You could design some sort of little see-saw probe that is spring loaded and tilts up and down with each heartbeat a little bit (maybe 0.1" or so?). Then you can come up with some sort of graphical recording mechanism like you described in your post.

Let us know what you find!

https://thumbs.dreamstime.com/z/fir...mative-medical-guides-first-aid-162185764.jpg

* The "anterior" side of the wrist is for the "standard anatomical position", which has the arms and hands rotated so that the palms of the hands are facing forward, toward the anterior side of the person.

 

[hutchphd]

So now I need to find a tube that can best carry waves. I haven't really read in depth about elasticity or waves, but if I had to guess I'd say that materials with a higher Young's modulus tend to carry sound waves better. However here I have another question: are sound waves and waves that make things 'jump' identical? I.e. if I have a medium that carries sound well, does it carry the 'jumping' part well? I've bought a section of a hose pipe but haven't gotten to try it out yet, but I don't have much hopes for this.

These are very good questions. There is another subtlety here, which @berkeman partially addresses. In addition to transmitting the sound well, your "tube" must "couple" well to both the chest wall and the indicator mechanism without reflecting the energy. This is generically called impedance matching and typically occurs when the speed of sound in path media changes. Really it is what the tiny bones in your middle ear (malleus, incus, stapes) do also.
I think you should take his suggestion one step further and think about a blood pressure cuff. In the transition pressure range 80<P<140 mmHg that cuff will make the mechanical pressure gauge needle bounce I think. Voila! ?

 

[berkeman]

I think you should take his suggestion one step further and think about a blood pressure cuff. In the transition pressure range 80<P<140 mmHg that cuff will make the mechanical pressure gauge needle bounce I think.

That's a good idea, and yes you can get a lot more focused energy out of the BP pressurized cuff than just with a simple radial pulse. The downside is that the cuff is pretty uncomfortable, especially for any extended period of time while compressed/tight.

 

[hutchphd]

Gangrene? I think the most signal to noise may be as you go through 100 mmHg but you should have a better knowledge (and your finger on the pulse as it were...)

 

[berkeman]

I think the most signal to noise may be as you go through 100 mmHg

Yeah, depending on the patient, you can get the most needle bounce on the pressure gauge a little above the Diastolic pressure, say 10-20mmHg over. So it won't be as uncomfortable as when the cuff is inflated above Systolic pressure (especially 20mmHg over like is done by some machines and some nurses...). I also have plenty of patients where there is little or no needle bounce on the pressure gauge while I'm taking their BP, but I can still hear the Korotkoff sounds in a quiet environment.

https://www.ncbi.nlm.nih.gov/books/NBK539778/

 

[DrBanana]

These are very good questions. There is another subtlety here, which @berkeman partially addresses. In addition to transmitting the sound well, your "tube" must "couple" well to both the chest wall and the indicator mechanism without reflecting the energy. This is generically called impedance matching and typically occurs when the speed of sound in path media changes. Really it is what the tiny bones in your middle ear (malleus, incus, stapes) do also.
I think you should take his suggestion one step further and think about a blood pressure cuff. In the transition pressure range 80<P<140 mmHg that cuff will make the mechanical pressure gauge needle bounce I think. Voila! ?

Can that needle provide enough torque (I think that's the term) to move a stick, or at least a pencil?

 

[hutchphd]

What needle? A cuff is a donut shaped bladder constrained by an outer ring. When inflated it uniformly squeezes your arm until the blood flow is restricted. The pressure then pulsates in the bladder..

EDIT: Apologies I used the term needle. Doh. So you good?

 

[berkeman]

Can that needle provide enough torque (I think that's the term) to move a stick, or at least a pencil?

Is there a reason that you want to use a Neanderthal approach to graphing this data? There are many better ways to do what you want. Is this a TikTok challenge or something?

What needle?

He initially posted that he wanted to make some Edison style stylus recording apparatus.

 

[hutchphd]

I think there is plenty of torque if used wisely. But other than for proof of concept @berkeman questions are mine also.

 

[BillTre]

Would a record player needle/cartridge work?

Could be hooked up to electronics.

 

[Tom.G]

The idea of using a pencil could be improved by using a fountain pen or a capillary pen (think of a very tiny straw), both have a lot less friction on the paper which means less energy needs to be collected to move them.

Or for experimenting, move a tiny mirror with a light beam shining on it. That way you can get a very long 'arm' to the screen or wall with practically no friction or moving mass.

Cheers,
Tom

 

[DrBanana]

Is there a reason that you want to use a Neanderthal approach to graphing this data? There are many better ways to do what you want. Is this a TikTok challenge or something?He initially posted that he wanted to make some Edison style stylus recording apparatus.

No particular reason other than I kind of like doing things in a more 'mechanical' way (no electronics). I don't use tiktok.
That being said I don't mind hearing what some of these other ways are. BillTre's idea seems promising but I don't know where to find a record player (plus it's gonna use electronics anyway).

 

[BillTre]

I don't know where to find a record player

They are not hard to find.
You can get them from many places: thrift stores, some electronics shops, or a web search will surely give you something.

 

[sophiecentaur]

No particular reason other than I kind of like doing things in a more 'mechanical' way (no electronics). I don't use tiktok.
That being said I don't mind hearing what some of these other ways are. BillTre's idea seems promising but I don't know where to find a record player (plus it's gonna use electronics anyway).

It turns it into a different sort of problem and could be very impressive. Actually, I reckon there's actually quite a lot of available 'signal power'; you can feel a pulse with a finger. I think it will be a matter of Matching the impedance of the blood pressure and displacement. I'd consider hydraulic transmission and matching, rather than a pneumatic system. I suspect that a diaphragm over the end of a cyinder held against the wrist (or, even better, the neck) would displace the water. A diaphragm could have an area of perhaps 1cm squared and the displacement would be 'just' visible so maybe 1mm. A rigid tube with a thin vertical (say <1mm squared) tube. The level of the water in the tube would be say ten times the displacement of the diaphragm.

Systolic blood pressure is around 0.1Bar, (I reckon) which is about 100mm of mercury and a range of a few tens of mm. That would correspond to ten times the height of a water column if you amplify the displacement with a suitable taper of tube bores. A float and lever could move a pen . . . . .

 

[.Scott]

If you are sticking to a mechanical recording device, you need to consider your bandwidth. A well-equipped engineering group might be able to chart 100Hz signals by entirely mechanical means. But as a home project, I would target about 10Hz - little more than the beating itself.

Given that limitation, any more-or-less inelastic tubing will be fine - as long as it isn't stretching in response to the signal. While the signal is being transmitted through air pressure, impedance coupling will be a minor issue because the wavelengths involved (>30m) will far exceed the size of your device.

The cuff is great and why not use the needle that comes with the mechanical sphygmomanometer (about $20) to drive the recording pen.

 

[BillTre]

BillTre's idea seems promising but I don't know where to find a record player (plus it's gonna use electronics anyway).

There used to be mechanical phonographs:

 

[sophiecentaur]

A pencil at the end of a stick or something will be writing to this piece of paper.

Way back, I used an ink on paper strip recorder but I reckon a sharpie would leave a line with much less drag than a pencil. The pens we used had a reservoir, a narrow tube and a sort of pen nib which was held close but not touching (iirc) the paper; very low drag - just surface tension.. But felt tips were not invented in those days.

The cuff is great

I think you could be right about the basic system. There's a lot of air shifted and i can feel it happen when I use (infrequently) my heart pressure monitor. I would still suggest a hydraulic system could do better than pneumatic. Of course you'd need to be careful about the static pressure but you wouldn't be looking for and stopping Systolic pressure level - just a signal that's proportional to the pressure - or actually displacement of blood to the arm. Some very flexible tube (bicycle inner tube?) could be wrapped around the arm and filled with water. Total displace water volume could be easy to see in a narrow U tube.

 

[OCR]

There used to be mechanical phonographs. . .

Lol. . . you forgot the dog.

.

 

[sophiecentaur]

Would a record player needle/cartridge work?
View attachment 323306
Could be hooked up to electronics.

Basically 'no-chance', I'm afraid. Th reason is that it works by magnetic induction in which the induced emf is proportional to rate of change of magnetic flux. (The time differential of the mechanical signal). It has incredibly poor low frequency response. A form of electrostatic transducer would have good DC response but you'd need what is essentially a variable capacitor in which the space varies with the displacement of the movement of a diaphragm (or whatever). But anything like that would be definitely not just mechanical.

 

[Tom.G]

Then there is the Crystal phono cartridge. The piezo-electric effect generates a nominal 1 Volt output when playing a record.

Drawback is similiar to a magnetic cartridge in that it generates a signal only while being deformed, no DC output or positional information is available.

 

[sophiecentaur]

The only way to achieve an electrical signal that's 'DC coupled' is to use a high frequency signal, applied to a transducer. The amplitude (or frequency) can be modulated by some dimension in the transducer and the steady state value (i.e. what would be DC) is transformed into a high frequency which can be detected with a much better signal to noise ratio.

Otoh, a mechanical quantity like pressure or displaced volume gives a very much DC coupled mechanical signal. The only problem in my mind is that I have never come across a pre-electronics recording sphygmomanometer. I'm sure the Victorians would have tried (and failed).

 

[Tom.G]

The only way to achieve an electrical signal that's 'DC coupled' is to use a high frequency signal, applied to a transducer.

I seem to recall than Alexander Graham Bell solved that problem by using a Carbon microphone.

In fact, they were in use in the rotary dial telephones and even early TouchTone phones, until phones started using active 'Electronics' instead of passive innards.

But the OP has stated that a mechanical approach is desired, so . . . moot point.

Cheers,
Tom

 

[sophiecentaur]

I seem to recall than Alexander Graham Bell solved that problem by using a Carbon microphone.

In fact, they were in use in the rotary dial telephones and even early TouchTone phones, until phones started using active 'Electronics' instead of passive innards.

But the OP has stated that a mechanical approach is desired, so . . . moot point.

Cheers,
Tom

Good thought; I like it. A variable resistor. You could use a cuff full of carbon grains? I wonder whether there's a source of them.

It is more of an 'electrical system' the electronics so maybe the OP would accept the idea.

 

[TonyStewart]

So my question is, is this feasible?

No.
Imagine you are completely motionless with an awkwardly balanced, contraption to avoid interference from muscle artifacts, lung & abdomen motion, laying down to avoid the imbalanced forces from gravity and a lot more challenges, quite simply, no. Everything induces failures and no pencil imprint control is possible.

The gain in function (force, displacement, power) is too many orders of magnitude to expect from a passive solution.

Consider the fact that most Western Doctors cannot even perform an acupuncture pulse diagnosis on the wrist with 2 or 3 fingers, or a palm on the abnomen or 2 fingers on the neck to feel the quality of palpations, because there is very little motion.

Yet, any $25 IR oximeter across any finger or toe can accurately plot the modulation of pressure by the transmission of light thru the network of blood expansion. Then display the pulse rate in BPM and O2 concentration using the ratio of IR & Red light in a matter of seconds. However, if your finger is also touching something causing vibrations in pressure, this will disturb the automatic gain (AGC) and offset (AOC) control of detected light to optimize the signal.

I would try to let go of your DIY Flintstone idea (but nice try) and get yourself a Pulse Oximeter and compare fingers, toes for O2 content from aging lung disease or COVID if it drops well below 90%. (then get help soon) If you can't get a pulse, seek help very soon. (it happened to me in my right median artery, with arm pain symptoms, then after 5 "roto-rooters" still no wrist pulse next day, followed by Median Artery bypass 3.5 yrs ago and still have neuropathy in 3 fingers related.

 

[berkeman]

Consider the fact that most Western Doctors cannot even perform an acupuncture pulse diagnosis on the wrist with 2 or 3 fingers, or a palm on the abnomen or 2 fingers on the neck to feel the quality of palpations, because there is very little motion.

What's "an acupuncture pulse diagnosis"? And what doctors are you seeing who are having trouble palpating an accurate radial or carotid pulse?

 

[TonyStewart]

What's "an acupuncture pulse diagnosis"? And what doctors are you seeing who are having trouble palpating an accurate radial or carotid pulse?

I'm not talking simply rate. I'm talking about a field of Asian medicine with 3k yrs of R&D with no contraindications. https://www.dc-acupuncture.com/physical-health/how-pulse-diagnosis-works

After I asked my Persian personal Dr. and Jewish Hematologist why after 3mos of novel 24x7 migraines from apparent anemia yet high ferritin they could not diagnose the problem after 3mos, yet Traditional Chinese Medicine (TCM) cured it in 1 to 2 weeks later ( after I gave up)

Their answers were; they don't teach TCM in Western Medicine. I replied, what is stopping you from learning every day, as do some Engineers (myself included)? (shrug)

 

[TonyStewart]

Added oximeter display.

%SpO2=91 %
P.I.% = 5.1 %. As the peaks get bigger, this increases
Rate = 68 BPM

 

[berkeman]

SpO2=91

That's pretty low. Is that normal for you?

 

[TonyStewart]

That's pretty low. Is that normal for you?

No. Thanks for asking. Yes anything consistently < 90% is of concern and 91% seems be the 1st time I've seen it this low too. Perhaps the low battery icon has something to do with accuracy, but I think it's just spending too much time resting on PF. I took another reading now and it averaged 95% sand went up to 97-98% momentarily when I scurried into the kitchen while breathing faster. But static breathing rate changes normally do not change %SpO2 as much as one expects and hardly budges it, indicating other chemistry chains need to be triggered for demand work. At age 70 in otherwise good condition it was consistently 98% last year when I was a "smoker" . Go figure, yet my lung spirometry tests always scored well when healthy. Red Vf being near 2V and close to a low battery voltage of two AAA's in series might explain the error but I would expect a better design from the Chinese but that might add a few pennies to the cost. ;)

 

[TonyStewart]

Aha it was the low battery tank warning causing the error.

I was just kidding about spending too much time "resting on PF".

or was I. ;) The %O2 error depends on the accuracy of the constant current regulator at low voltage and how that affects the amplitude and ratios of IR / R or is it R/IR, light detected. I forget.

 

[berkeman]

An off-topic discussion about Traditional Chinese Medicine has been deleted. Since we have addressed the OP's questions and they haven't been back in several weeks, the thread will be closed now. Thank you to all for your contributions and ideas for the OP.