engineer effort to help COVID-19 testing: low cost real-time PCR

iammegatron

In the same spirit as the calling for ventilators, I am thinking about building low cost real-time PCR (qPCR) machines that can help COVID-19 testing.

Propeller 2 can certainly perform all the electronics tasks, such as driving Peltier for thermal cycling, lighting LEDs and reading CCD sensors etc.

The main obstacle would be the optics assembly design and florescence probes (I heard some of the chemicals are toxic and out of the reach of EE engineers).

Any discussion of methods, vendors etc.

ManAtWork

Basically, good idea. I suggested that in another forum a month ago and also got zero response. I don't understand it but most engineers only think about building ventilators. IMHO, more tests would save more lifes than more ventilators.

Patients who need a ventilator require ~10 days treatment and have a very bad chance to survive, <50% if I remember correctly. So 1 ventilator saves 1 life per 20 days.

If you have an automated PCR tester that can test 100 samples per day... Lets assume 10% of the tests are positive and the reproduction rate is near 1. So by discovering 10 possible infection paths (and of course avoiding them) per day you'd save 1 life per 10 days if mortality is 1%.

ManAtWork

Ok... this simple calculation is probably not as convncing as I first thought. If you don't have enough ventilators more people die, so mortality would be probably 2% instead of 1% and the game ventilators vs. testers ends undecided.

But not in the long run! If you have considerably more than one tester and test really a LOT then the reproduction ratio is likely to drop. If you only invest in ventilators you can treat people who are already infected but you can do nothing against the spreading of the virus. Reproduction rate is likely to raise.

Prevention is better than treatment.