Something like an air monitor which which can be installed in a classroom which can detect the COVID virus (and others), and alert the teacher who can take action before it gets a chance to spread? Or perhaps another form like a breath analyzer?
If such a device can exist it may be more effective then a vaccine.
I think i read about dogs being trained to do that, with a high success rate. I don’t think that will catch on, though.
Haven’t they trained dogs to do such things? That seems a lot more likely than what your suggesting, to be honest. Probably CAN be trained, and could then, quite quickly, scan a whole classroom of kids.
I think it’s maybe possible.
(Ninja’d!)
I doubt a dog could do it, at that level of specificity. A dog might be able to smell that someone’s been sneezing or coughing in the room, but even their noses aren’t sensitive enough to specifically identify one particular virus.
With a device, it might be possible, but only for a specific disease or a short list of them, and only if virus-containing particles happen to land on the device. And if they’re landing on the device, then they’re also already landing on people.
Does it have to ‘land’ on the device, would it be possible to detect a EM signature of the molecule by perhaps hitting it with some light wave?
You’d be competing with the EM signature of literally every other molecule in the room. No way that’s feasible.
The absolute easiest you could do would be to stick a probe up everyone’s nose as they enter the room and do a rapid test on each of them. That, we could do right now, except that it’s expensive and annoying.
https://www.nature.com/articles/d41586-020-03149-9
I don’t know, this article makes it seem possible and promising.
Most of this is Star Trek level of science fiction.
A COVID-19 virus is a strand of RNA of 29,811 base pairs within a ball of lipids with a whole host of specialised (ie the spike) proteins on the surface. The spike protein is 1273 amino acids long.
Nothing here has something as simple as a specific signature. We work out what they are by laborious (albeit highly automated) laboratory analysis.
All of life is filled with DNA, RNA, and huge numbers of different proteins, all built from the same building blocks. About all we could determine by zapping a droplet of gunk in the air would be a rough approximation of the elements that it contained. Given that would be pretty much guaranteed to be oxygen, hydrogen, nitrogen, carbon, plus you might just detect some residual other stuff, enough to say that well, maybe, whatever it was, was probably once part of something that was in some form alive. At a bleeding edge stretch you might detect some amino acids. Still not much of an improvement.
Rapid tests for a specific protein is possible by creating antigens that bind to that protein and using some form of labelling on the antigen to work out when it is binding to stuff. This is a one trick test. You make it for the one target. But you need to get the target onto the antigen to get it to bind. And you need more than the one particle, so swabs are the order of the day.
Smells are really interesting, with perhaps the scent of Parkinson’s Disease being one of the more interesting. But these are not the scents of the pathogen, but the scents of the body’s response, or even the scents of body flora products changing in response to the body’s response.
It’s quite common in medical research for preliminary results to look promising, but those promising early results almost never pan out.
Here, I suspect that the biggest issue is lack of proper control. The results they’re claiming could just be dogs distinguishing between people with respiratory disease and people without respiratory disease. How well do they do in distinguishing between people with covid, the flu, the common cold, and seasonal allergies?
define disease. there are something like 60,000 things that can go wrong with the human body, and the types of diagnostic tests used to diagnose them vary wildly.
Then you have issues of false positives and false negatives.
There is research into attaching various medical diagnostic devices to cell phones, so that medical diagnosis can be done more remotely. cell phones can be used as ekgs, ultrasounds, to diagnose malaria, etc. that’ll be big in the next few decades to bring medical care to poor and isolated regions of the planet.
I had flu a lot as a teenager. And once i got the weakened virus nasal flu vaccine. And I recognized that smell. It smelled like being sick with the flu.
I’m told that back before modern testing, doctors used to diagnose many illnesses by their characteristic odor.
I think it’s extremely likely that a dog can recognize what a person sick with covid smells like, and reliably distinguish between covid and other common respiratory bugs. Dogs have a vastly better sense of smell than i have, and i can smell flu, at least in my own nose.
I don’t think it will catch on, for a variety of reasons. Trained dogs aren’t uniform in the way machines are. They are expensive to maintain, and you can’t just ship them around, they need a social connection to their handler. At best, maybe we’ll use them in airports.
But i completely believe that a dog can scan people quickly for covid.
Dogs and their ability to detect disease was the subject of Malcolm Gladwell’s latest Revisionist History podcast.
There is an oddly specific smell/taste combo in the way back of the nose and throat [for me at least] when they would do that first push of saline into my port, there is a distinct smell/taste way in the back of my nose and throat when I get into the presence of phosgene [like a fresh mown lawn with that hand full of change in a hot sweaty hand deal] so I can see using dogs to diagnose certain things by scent.
That would be the lab report I guess. They only do that if the cat scan is unsuccessful. 
One of the problems with any sort of fast and mass applied test (be it a sci-fi or trained dog) is the details of how it is intended to be used and how good the test is.
No test is 100% perfect, and there is devil in the details.
The false positive rate and false negative rates make a big difference to its utility, and affect how you can usefully deploy the test.
If your test has a 1% false positive rate and the disease is rare, say 0.1% of the general populate has it at any one time, the test may be useless. It will turn up 9 people without the disease for every one with it. If you need to find everyone with the disease no matter what, this might be acceptable, you pull them out of circulation and do further tests. But in a classroom setting for everyday use, you may find it impossible to operate with a constant barrage of false alerts.
Similarly, false negative rates are important when you need to allow people to get on with their lives. If the disease is endemic, you need a very low false negative rate in order to be allowing people to go back to their ordinary lives, otherwise you can’t usefully contain things.
There is no hard and fast answer, the utility of a test depends upon the expected fraction of the populace being tested that has the disease at any time. And it thus depends upon the use expected of the test. There is a big difference between testing everybody in a classroom or street, versus testing a patient that presents with clear symptoms and you are trying to exclude or include specific pathogens.
I see all the things you did there.
There are indeed breathalyzers in trials right now, but they detect common chemical signatures of the body fighting particular kinds of pathogen. They don’t directly detect the virus.
It’s true that the latter would be the holy grail for fighting infectious disease, but I don’t think there is any sign of a technology that would come close to achieving this in real time.
That said, the “chemical signatures” style breathalyzer would be a game-changer for covid if accurate, regardless of underlying mechanism.
Often, not only are there different tests with varying rates of false positives or false negatives, but they can even be the same test, with a different threshold set. Maybe what the test actually determines is a “covid index”, a number that ranges from 0 to 10. And you know from your research that a person with a covid index of 2 absolutely never has the disease, and a person with an index of 8 absolutely always has the disease. But 1 time in 100, someone with the disease might have an index as low as 3, and 1 time in 100, someone without the disease might have as high as a 7, and so on.
But you want a test that just returns a yes or no. I gave the test; should this person be allowed into the building? Depending on how common the disease is and how bad the consequences are for each kind of error, you might set different thresholds. If you want to be absolutely certain that you don’t let anyone at all in with Covid, then you only let people in if they’re 2 or lower, but the cost of that is that you’re excluding a lot of healthy people needlessly.