Slashdot Mirror
More Medical Devices Should Be Open Source, Like This ECG (github.com)
isza writes: This is a follow-up to the Slashdot story about mobilECG, a 12-lead, clinical-grade ECG being open sourced. We have not given up on our goal to get rid of the high-profit-margin and dishonest distribution practices of diagnostic ECGs, and make a certified open source version of this important diagnostic device. After many months of hard work, there is now a working prototype of a much more capable device than the first version, with its sources available on GitHub. MobilECG now has a Holter function, changeable lead-configurations and Bluetooth. Here's a video of the prototype working.
"They are custom-fitted, high precision medical devices..."
Which contain perhaps a hundred bucks worth of electronics and have a history of seldom working to the satisfaction of the patient. My town is full of high-end retirees, and every one I have spoken to has a drawerful of expensive hearing aids that suck. This area of medicine is overripe for disruption by some Silicon Valley company that can make a device that performs more like a natural ear.
You will find that the analog front end for designing an EEG is non-trivial. Humans are a very low impedance source with a very poor signal to noise ratio due to cable, electrode, muscle and other subject generated noise that swamps the signal. Large sacks of electrolytes make great antennas for 50/60 Hz pick-up too. Add in the ability to survive a defibrillator hit of a life-safety device and the analog front end become very interesting. That said, TI has an entire analog front end for ECG that handles many of the thornier challenges including actively driving the leg electrode.