I am a biologist with a little (very little!) experience fiddling with electronics and arduinos looking for help with a project/problem we have. My experience is limited to building a 3d bioprinter and using an arduino to power a little peristaltic pump.
I work on tissue engineering projects and recently we have had some success making little cardiac tissues. They are made up of stem cell derived cells including those that beat (called cardiomyocytes). We are doing a bunch of experiments to see how well they can be used as a disease model, but one of the outputs we need is electrical activity. Essentially when artificial cardiomyocytes (and nerve cells) are made one of the best ways to validate them and use them to study disease is to monitor their electrical activity with a device called an MEA (multielectrode array - examples here: https://www.multichannelsystems.com/products/microelectrode-arrays
Unfortunately I am quite a junior academic (read as essentially broke) so I am looking for a cheap alternative to commercial MEAs (which are frighteningly expensive). We found and reassembled an abandoned MEA which is 20 years old, but it is a bit of a mess and half the electrodes don't work. Our other problem is that normally a flat 2D monolayer of cells are seeded onto electrodes in an MEA, in our case we have a round floating blob of cells about 2mm across in size.
I did some googling and reading and it seems like there are a good number of ECG/EMG projects based on Arduinos. What I was hoping for was some guidance about whether it would be feasible to do something like that, perhaps linked to a chip like this one: https://ibidi.com/impedance-measurement/164-ecis-cultureware-8wcp-pet.html
Ultimately what we need is something which allows us to:
Measure the electrical activity of cells in a well in a cell culture slide (like the Ibidi one I've linked to above) for 2-10 minutes.
In a 37 degree environment - we have incubators set to this temperature.
Ideally we would also be able to programme a 'pacing' protocol whereby we do baseline readings for a set amount of time, then move on to stimulate the electrodes at a particular frequency (e.g. 1Hz) and record that etc.
Would anyone be able to advise on if this is feasible? If it is wiser to pack this in I will do so!
And if feasible any suggestions on the simplest way to go about it? I have a couple of arduino unos, and I had noticed there are products like the MyoWare muscle sensor which could help/work.
Any thoughts and suggestions very welcome. Happy to provide more detail if needed.
You will likely discover why such equipment is expensive.
But, you already know the voltages you are dealing with. Now you need to discover the voltages necessary to be detected and measured by your Arduino. Now you know how much amplification is necessary for your project. Concentrate on building or buying a DC amplifier that will do the job, and not introduce any stray signals.
Paul
I am sorry your link took me to a page where I had a choice of several products they make, I had no idea of what to click. I would assume the low level amplifiers are very expensive and hard to come by. I did not find where you stated what voltage you will be measuring. If possible get to know some analog dude, the help will be invaluable. I know nothing about what you are doing but I am assuming you are in the nanovolt to microvolt range, probably AC type of signals. Sorry I cannot help more.
Thanks for taking the time to respond! If it helps basically the products they make are all variations on the theme of detecting cell signals - they vary the format (sometimes we grow cells on chips and sometimes on multiwell plates).
Re voltage range - looking at the literature we expect signals between 50uV and 1500uV depending on the sample. As we haven't measured these ourselves (due to lack of kit) we think it will sit in there based on the literature. I think it would be an AC signal, but I'm not really sure I will dig into it.
That gives a native resolution of 1.1/1024 ie about 1mV
and with a gain of 1000 a resolution of 1uV.
I'm not aware of ANY biological system that produces DC.
Thank you! That's very helpful. Would a filter be needed along with this? Sorry if this is a painfully stupid question, just wondering about how much interference/noise will be an issue,
you can implement a digital filter in arduino.
Until you try it you wont know what the sources and levels of interference are.
Having tried it you could move from the arduino's native ADC to using a better ADC.
Read this thread regarding noise decoupliong and grounding
Coincidentally my testbook on DSP, "Digital Signal Processing a Practical Approach", Ifeachor&Jervis,
has biomedical signal processing as worked examples including EEG and ECG. From what I remember
you'd normally use a high order phase-linear FIR bandpass filter on such signals to knock out mains
frequencies and above, and DC drifts. Mains frequency and its harmonics are a major source of
interference.
I think its quite a lot of work to get high quality signals as you'd need good analog anti-alias filter,
then sampling around several 100 Hz, then perhaps 100+ tap FIR lowpass filter with cutoff 40Hz or so,
an IIR filter to lose DC drift, and all for many channels simultaneously?
Hi Mark - I used to work as an electroencephalographer.
EEG signals are much smaller than ECG, and muscle EMG was also an issue. We certainly in those days didnt have ANY advanced filtering. The LF signals were chnged to AC with electromechanical choppers!
However in the scenario described these are not present, and i believe other sources of interference can easily be minimised.