MIT researchers develop an AI model that can detect Covid-19 in asymptomatic individuals

Status
Not open for further replies.

Polycount

Posts: 3,017   +590
Staff
In brief: Though global economies have begun to open up, the Covid-19 virus is still spreading throughout the world, infecting thousands of new people every day. To help curb the spread of the disease, MIT researchers have developed an AI model that can detect the virus' presence in even asymptomatic individuals.

The potential good that such a model could do is probably pretty obvious. Suppose the model was refined and rolled out to the general public somehow, perhaps in the form of a free mobile app.

In that case, it could help people screen themselves for the infection and either get tested or avoid contact with others, if necessary.

Teachers, for example, could use it every day before heading into class, as could other individuals that work in close proximity to strangers -- front-line retail employees are another key audience for such a tool.

Fortunately, porting the model to an app is precisely what researchers are working toward now. The team still needs to finish developing it and, of course, obtain FDA approval before it can be distributed widely, though.

However, early results are promising. To date, the model has been trained on "tens of thousands" of cough samples, as well as "spoken words." When researchers input new cough recordings into the model, it can accurately identify full-fledged Covid-19 infections in "98.5 percent of coughs" and 100 percent for asymptomatic individuals.

It seems that, even for people who aren't showing severe symptoms of Covid-19, the way they cough, breathe, and speak may contain small indicators that can point toward probable infection.

If this model truly is as accurate as researchers believe, we hope the FDA approves it sooner rather than later -- the world could certainly use it right about now.

Image credit: Josep Suria

Permalink to story.

 
Gonna have to change the definition of asymptomatic soon. Glad to see this innovation - but since so many people seem to think “Covid is a government hoax” or “it isn’t really that bad”, I worry that it might not accomplish much.

Was watching college football yesterday and was actually physically nauseated to see tens of thousands of ID1OTS in the stands...
 
Was watching college football yesterday and was actually physically nauseated to see tens of thousands of ID1OTS in the stands...
Why? I imagine the great mass of those individuals were neither elderly or had severe comorbidity factors, meaning they were less at risk to Covid than they'd be from the common flu. They were also outdoors, I assume, which cuts the risk factor down by another order of magnitude.
 
Gonna have to change the definition of asymptomatic soon. Glad to see this innovation - but since so many people seem to think “Covid is a government hoax” or “it isn’t really that bad”, I worry that it might not accomplish much.

Was watching college football yesterday and was actually physically nauseated to see tens of thousands of ID1OTS in the stands...

College students are almost immune to Coronavirus. Their CFR is extremely low. Stanford came out with a study recently showing that for healthy people under 65, the chance of dying is the same a daily auto commute to work.

The big risk is for those over 75, with at least one comorbidity. Those are the ones who should self isolate. Everyone else should have a choice.
 
College students are almost immune to Coronavirus. Their CFR is extremely low. Stanford came out with a study recently showing that for healthy people under 65, the chance of dying is the same a daily auto commute to work.

Go ahead and link to that study and its conclusions. You may find that the study actually made a different claim which was twisted by groups with a specific agenda to suit their belief system.
 
Why? I imagine the great mass of those individuals were neither elderly or had severe comorbidity factors, meaning they were less at risk to Covid than they'd be from the common flu. They were also outdoors, I assume, which cuts the risk factor down by another order of magnitude.

Not so. If you're 15 or older, your risk of death from covid is greater than from the flu:

https://freopp.org/comparing-the-risk-of-death-from-covid-19-vs-influenza-by-age-d33a1c76c198

And remember, that's with the social distancing, masks, and other preventive measures in place which the flu numbers don't benefit from. Without those, covid deaths would be even worse.
 
Why? I imagine the great mass of those individuals were neither elderly or had severe comorbidity factors, meaning they were less at risk to Covid than they'd be from the common flu. They were also outdoors, I assume, which cuts the risk factor down by another order of magnitude.
The stands were FULL in many of those games... and it doesn’t matter if they were elderly or not.... these id1ots were exposed - and will certainly spread it to others...

The argument of “if I get Covid I’ll be fine” or “I’m ok with the risk of getting Covid” is what’s killing thousands of high risk people... because these m0rons then infect others!!

Even if the risk is “low” - when 60-70,000 people congregate in one location, the risk becomes HIGH!!

And very few of the fans were wearing masks either - not that that would have helped much being that close to others....
 
College students are almost immune to Coronavirus. Their CFR is extremely low. Stanford came out with a study recently showing that for healthy people under 65, the chance of dying is the same a daily auto commute to work.

The big risk is for those over 75, with at least one comorbidity. Those are the ones who should self isolate. Everyone else should have a choice.
The risk of dying might be low - but the risk of infection is the same.... and now these people will infect others!

There should NOT be a choice!!

Be responsible and try not to kill your fellow Americans!
 
Why? I imagine the great mass of those individuals were neither elderly or had severe comorbidity factors, meaning they were less at risk to Covid than they'd be from the common flu. They were also outdoors, I assume, which cuts the risk factor down by another order of magnitude.

Possibly true. But they may pass it on to others who are not young or healthy (I assume they have families).

People always forget this part.
 
Not so. If you're 15 or older, your risk of death from covid is greater than from the flu: (link)

Not true. Your link is based on data from February-September. Survival rates from Covid have improved dramatically since the first few months of the pandemic, which explains why we are now generating more than twice the new cases/day as we were in April, yet death rates are less than half what they were then. Survival rates for the flu, however, have of course not changed.

And remember, that's with the social distancing, masks, and other preventive measures in place which the flu numbers don't benefit from.
Eh? Social distancing reduces flu transmission just like Covid. As for cloth masks, much research demonstrates that they are ineffective and even dangerous. An N95 respirator or even a 3-layer cotton mask, worn indoors by people trained on how to properly wear, handle, and dispose of it does provide some small benefit. But a thin piece of cloth over your face, worn and handled in the manner in which nearly all the general public does, increases transmission risk through at least three separate factors. I'd be happy to provide you links to peer-reviewed research demonstrating this.

Possibly true. But they may pass it on to others who are not young or healthy
A statement true of the flu and of all infectious diseases. We could in fact eradicate every infectious disease known to man -- diseases which cumulatively kill far more individuals than Covid -- if we simply stopped all human interaction. However, we're smart enough to realize that the negative aspects of such a course far outweigh the benefits. Or, I should say most of us are smart enough to understand that.
 
Last edited:
As for cloth masks, much research demonstrates that they are ineffective and even dangerous. An N95 respirator or even a 3-layer cotton mask, worn indoors by people trained on how to properly wear, handle, and dispose of it does provide some small benefit. But a thin piece of cloth over your face, worn and handled in the manner in which nearly all the general public does, increases transmission risk through at least three separate factors. I'd be happy to provide you links to peer-reviewed research demonstrating this.
No, that research has been debunked... I assume you're talking about the "findings" from this article.

Well, here's the flaw in that article - I've taken this from Dr. Andrew Morris - you really should subscribe to his newsletter, as he is one of the leading experts...
_______________
First, we know that aerosols are not the major source of transmission, because most infected people have been in close contact with an infected person. (If the major source of transmission were aerosols, we would be seeing plenty of infected people who were never in close contact with a case.) The CDC, WHO, and others have acknowledged that aerosol transmission occurs and is important, but they also point out that the main route of transmission is "droplet" (I.e. ballistic or close contact).

So what did the authors do? First, they start off with this: At the beginning of the pandemic, it was believed that the large droplets we expel when we cough or sneeze were the main vehicle of transmission. This sets the reader up to believe that this initial idea was wrong. There is no evidence it was. They then go on to say an article in the prestigious Science magazine found that there is “overwhelming evidence” that airborne transmission is a “major transmission route” for the coronavirus. Except the article wasn't an article "finding" overwhelming evidence, but rather it was a letter claiming overwhelming evidence. Two top US scientists, Drs. Angela Rasmussen (virologist) and Saskia Popescu (epidemiologist) told me they refused to co-sign the letter because of the misleading wording.

Finally, the beautiful visualizations—and they are beautiful—lead the non-expert reader to believe that they approach some kind of truth, even though they acknowledge some of the limitations. I guess the visualizations are what attracts readers. But the simulations used for these graphics are based on models that are entirely unrealistic. For example, we know from many studies that household members have, at most, a 50% chance of getting infected from another member. (That high number comes from a study that came out today, and may be lower overall.) And yet the model shows that there is a near certainty that everyone sharing a room with an infected person will be infected after 4 hours. It is an absurd model.

The last thing I will say on this article: it is a fantastically written and illustrated model, with many important takeaways. In fact, I agree on masking, and the benefits of ventilation/fresh air. But you will note that it doesn't even mention physical distancing. And herein lies the danger of over-emphasizing aerosol transmission: it seems that some experts keep on banging something so hard that they are even starting to drown out some of the most importance stuff. It is as if all they want is more cowbell.
_________________

Anyways... please, WEAR A MASK, don't spread misinformation, and be SAFE out there!!

Edit: oh, and check out https://covidemails.com/ and subscribe - you'll get REAL information about Covid, instead of misinformation, and you'll begin to understand why so many people are getting sick and dying...
 
No, that research has been debunked... I assume you're talking about the "findings" from this article.
No. You're quoting blog posts and news articles. I'm talking about peer-reviewed research published by the NIH, the CDC, and many other sources. Here are just three examples

CDC: Effectiveness of Cloth Masks for Protection Against Severe Acute Respiratory Syndrome Coronavirus

"We conducted a randomized controlled trial to compare the efficacy of cloth masks with that of medical masks and controls...Rates of infection were consistently higher among those in the cloth mask group than in the medical mask and control groups. This finding suggests that risk for infection was higher for those wearing cloth masks."

(note: control group wore no masks)

National Institutes of Health: A cluster randomised trial of cloth masks
"The study...results caution against the use of cloth masks. This is an important finding to inform occupational health and safety. Moisture retention, reuse of cloth masks and poor filtration may result in increased risk of infection....as a precautionary measure, cloth masks should not be recommended for HCWs..."

CDC : Emerging Infectious Diseases Journal
"In pooled analysis, we found no significant reduction in influenza transmission with the use of face masks... None of the household studies reported a significant reduction in secondary laboratory-confirmed influenza virus infections in the face mask group... Improper [mask] use might increase the risk for transmission... "(link
 
A statement true of the flu and of all infectious diseases. We could in fact eradicate every infectious disease known to man -- diseases which cumulatively kill far more individuals than Covid -- if we simply stopped all human interaction
As usual: veering off on a tangent (flu vs Covid-19) and setting up straw-man extremes ("if we stopped all human interaction") to deflect from the main point, which is that mass gatherings create mass infection which then spreads further into society. Promoting the latter shouldn't be a choice. THAT'S the point.

The only difficulty is for those committed to neo-con / alt-right ideology to acknowledge common sense - in this and many other issues. So we get another dancing lesson.
 
No. You're quoting blog posts and news articles. I'm talking about peer-reviewed research published by the NIH, the CDC, and many other sources. Here are just three examples
CDC: Effectiveness of Cloth Masks for Protection Against Severe Acute Respiratory Syndrome Coronavirus
"We conducted a randomized controlled trial to compare the efficacy of cloth masks with that of medical masks and controls...Rates of infection were consistently higher among those in the cloth mask group than in the medical mask and control groups. This finding suggests that risk for infection was higher for those wearing cloth masks."

(note: control group wore no masks)

National Institutes of Health: A cluster randomised trial of cloth masks
"The study...results caution against the use of cloth masks. This is an important finding to inform occupational health and safety. Moisture retention, reuse of cloth masks and poor filtration may result in increased risk of infection....as a precautionary measure, cloth masks should not be recommended for HCWs..."

CDC : Emerging Infectious Diseases Journal
"In pooled analysis, we found no significant reduction in influenza transmission with the use of face masks... None of the household studies reported a significant reduction in secondary laboratory-confirmed influenza virus infections in the face mask group... Improper [mask] use might increase the risk for transmission... "(link
Well, here's the abstract from your first article:
Cloth masks have been used in healthcare and community settings to protect the wearer from respiratory infections. The use of cloth masks during the coronavirus disease (COVID-19) pandemic is under debate. The filtration effectiveness of cloth masks is generally lower than that of medical masks and respirators; however, cloth masks may provide some protection if well designed and used correctly. Multilayer cloth masks, designed to fit around the face and made of water-resistant fabric with a high number of threads and finer weave, may provide reasonable protection. Until a cloth mask design is proven to be equally effective as a medical or N95 mask, wearing cloth masks should not be mandated for healthcare workers. In community settings, however, cloth masks may be used to prevent community spread of infections by sick or asymptomatically infected persons, and the public should be educated about their correct use.

Your second article is from 2015 and therefore irrelevant since Covid 19 didn't exist yet...

Your 3rd article is about Influenza - and therefore is also irrelevant.

No one is arguing that medical masks are better than cloth - but cloth is still better than nothing!

America is the land of the free, not the land of the oppressive and over reaching government. If there is a time those in government think they can control the citizens, oh do they have a lesson in history waiting for them.
Perhaps... but your freedom has always ended whenever it interferes with someone else's freedom!!

Example: You are NOT free to murder someone... as that takes away that person's freedom to live...
 
Not true. Your link is based on data from February-September. Survival rates from Covid have improved dramatically since the first few months of the pandemic, which explains why we are now generating more than twice the new cases/day as we were in April, yet death rates are less than half what they were then. Survival rates for the flu, however, have of course not changed.

I linked to data, you are making a claim. Show me the data.

Eh? Social distancing reduces flu transmission just like Covid.

Yes, but you missed the point. Flu data are from previous years where there was no social distancing & masks. Covid data is from a year with social distancing & masks. This means that if the flu data had been taken from a year with social distancing and masks in place, it was be far lower than those data I linked to.

In fact, that was the case late this year. This past winter's flu season was among the worst in recent years and was stopped dead in it's tracks when covid-based social distancing and masks came into effect. Note how quickly the cases drop to zero compared to previous years in March:

https://www.advisory.com/daily-briefing/2020/05/05/flu-update
 
Well, here's the abstract from your first article:
Cloth masks have been used in healthcare and community settings to protect the wearer from respiratory infections. The use of cloth masks during the coronavirus disease (COVID-19) pandemic is under debate. The filtration effectiveness of cloth masks is generally lower than that of medical masks and respirators; however, cloth masks may provide some protection if well designed and used correctly. Multilayer cloth masks, designed to fit around the face and made of water-resistant fabric with a high number of threads and finer weave, may provide reasonable protection. Until a cloth mask design is proven to be equally effective as a medical or N95 mask, wearing cloth masks should not be mandated for healthcare workers. In community settings, however, cloth masks may be used to prevent community spread of infections by sick or asymptomatically infected persons, and the public should be educated about their correct use.

Your second article is from 2015 and therefore irrelevant since Covid 19 didn't exist yet...

Your 3rd article is about Influenza - and therefore is also irrelevant.

No one is arguing that medical masks are better than cloth - but cloth is still better than nothing!


Perhaps... but your freedom has always ended whenever it interferes with someone else's freedom!!

Example: You are NOT free to murder someone... as that takes away that person's freedom to live...

You just equated exercising the right to live your life without a mask on as premeditated murder? *face palm*. How can anything you say be taken seriously at this point? Wow.
 
You just equated exercising the right to live your life without a mask on as premeditated murder? *face palm*. How can anything you say be taken seriously at this point? Wow.
No... I just gave an example of how being "free" doesn't mean you can do anything you want...
But honestly, going into a crowded place without a mask or social distancing WILL end up with more people getting sick or dying... it might not qualify as murder - but shame on you who are doing it!
 
No. You're quoting blog posts and news articles. I'm talking about peer-reviewed research published by the NIH, the CDC, and many other sources. Here are just three examples
CDC: Effectiveness of Cloth Masks for Protection Against Severe Acute Respiratory Syndrome Coronavirus
"We conducted a randomized controlled trial to compare the efficacy of cloth masks with that of medical masks and controls...Rates of infection were consistently higher among those in the cloth mask group than in the medical mask and control groups. This finding suggests that risk for infection was higher for those wearing cloth masks."

(note: control group wore no masks)

Those tests and recommendations were specifically for hospital healthcare workers, not for everyday use.

And your note about the control group is wrong. From the authors:

"It is important to note that some subjects in the control arm wore surgical masks, which could explain why cloth masks performed poorly compared to the control group."

Simply put, that is a *crappy* control and in fact is not a control at all. However it may be hard to get healthcare workers in Vietnam, or anywhere, to go without mask protection for long enough to do this type of study for a proper control.

So none of the conclusions from this study 5 years ago are necessarily applicable to mask wearing for covid in a social setting.

National Institutes of Health: A cluster randomized trial of cloth masks
"The study...results caution against the use of cloth masks. This is an important finding to inform occupational health and safety. Moisture retention, reuse of cloth masks and poor filtration may result in increased risk of infection....as a precautionary measure, cloth masks should not be recommended for HCWs..."

CDC : Emerging Infectious Diseases Journal
"In pooled analysis, we found no significant reduction in influenza transmission with the use of face masks... None of the household studies reported a significant reduction in secondary laboratory-confirmed influenza virus infections in the face mask group... Improper [mask] use might increase the risk for transmission... "(link

From the authors, in the Face Mask section:

"Most studies were underpowered because of limited sample size, and some studies also reported suboptimal adherence in the face mask group."

So the data is not very good.

This is the problem of a meta-analysis like this, the authors did no scientific tests of their own but instead looked at a bunch of other papers. This isn't necessarily a bad thing, but this means their conclusions are heavily biased on:

* how comparable any of the original studies are (this is usually a big problem as the individual studies are for different purposes)
* how good is any of the original science
* which studies they included and which ones they excluded

Any one of which can call a meta-analysis's conclusions into question, like the authors mentioned.
 
Last edited:
As usual: veering off on a tangent (flu vs Covid-19)
That's not a tangent, it's the heart of the matter. The question is what constitutes an acceptable risk? When compared to all the other risk factors of premature death, Covid is a small incremental increase, and in no way justifies the exhorbitant negative impacts of shutdowns, impacts which have not only destroyed livelihoods, but people's mental and emotional well-being also. Suicide rates are way up, homicides and violent crime are up, depression and other mental disease, divorce, spouse abuse, etc, etc. And don't forget an entire lost year for tens of millions of children, many of which will bear the developmental impact for the rest of their lives.

Promoting the latter shouldn't be a choice.
Wow. You not only want to illegalize gatherings, but illegalize the mere mention of it. Damn that pesky first amendment and our constitutional liberties, eh? Does the phrase "thought police" mean anything to you, Mr. Orwell?
 
Everyone can stay in their houses, and hide under their beds, until no one dies anymore! Sure, that will work. I mean, we have reams of data on how well the lockdowns have worked in Italy, Spain, the UK, etc back in April, May, June, and whenever else it was tried. Hint, it didn't work then, it ain't working now.

The case fatality rate for persons 19 and under in Canada is .0066%. Two, yes, two, people have died with covid out of about 30,000 positive "cases".

From 20 - 29 years of age 10 Canadians have died with covid, out of 42,000 positive "cases" for a CFR of .023%.

From 30 - 39 years of age 17 Canadians have died with covid, out of 35,000 positive "cases" for a CFR of .048%.

Those CFRs are far below the Infection fatality rate of the flu. In case you don't know, the CFR is always higher than the IFR because we just don't know for sure how many people actually have had this virus affect them.

In Canada about 95% of fatalities have been over 60, 9,620 of 10,055. Most of these folks lived in care homes, assisted living, or other types of "old folks" homes. I wonder what other flu seasons was like in these settings? Similar? Better? Worse? One thing is for sure, we really didn't keep track as closely as we are now.

https://health-infobase.canada.ca/covid-19/epidemiological-summary-covid-19-cases.html

I'm almost 60, overweight, but otherwise no other co-morbidities that I am aware of. My Mom is 80, she has a few of the things that might push her over the edge should she come down with ANY respiratory illness. My Dad passed away at 73 from pneumonia. I am familiar with the possibility of dying should I or a loved on get sick. But, for God's sakes people, no one lives forever. The reaction to this has been so out of touch with reality that it near makes me weep when I read the reactions here and elsewhere.



So, wear a mask, you can see how well it works by the sheer number of "cases" we have had since the start of September.
 
I linked to data, you are making a claim. Show me the data.
Are you seriously attempting to deny that mortality rates have improved dramatically? Besides the countless stories which have appeared in print on the subject, I refer you to any public Covid dataset, either from the CDC, or this somewhat more visual one. Daily case counts in April were averaging just over 30,000, and daily deaths nearly 2,500. Over the past 30 days, however, daily case counts were roughly 80,000 (one day hit 100,000), and yet the daily death toll just under 1,000.

That's nearly 3X times the cases translating into less than 1/2 the deaths. Obviously a more accurate methodology is to translate the case-count window back by a ~two week offset to better match cases to deaths, but this doesn't change the overall results. The results show a dramatic increase in survivability since the early days of the epidemic. Using current mortality risk rates, all age cohorts under age 45 have a lower risk of death from Covid than they do the flu, and you have to reach the 55+ cohort before the Covid risk becomes significantly higher.

Even worse for your argument is the fact that, even if we don't correct the data in this manner and use the inflated figures, they show that the age 15-24 cohort has only a 1.98X greater risk of Covid than of the flu. Thus a college student who in years past attended two stadium football games had the exact same risk of dying as a student today who attends one. In the first case you advocate saying and doing nothing whatsoever, but in second case you advocate banning the practice entirely (and, according to the prior poster, arresting anyone who even suggests attending a football game.) Do you honestly believe this is this a reasonable position?

This means that if the flu data had been taken from a year with social distancing and masks in place, it was be far lower[/quote]Checking your study, it apparently used flu data from the 2017 year. Not a terrible flu season by any means. Also, I reiterate that not one single study has ever shown public mask use to reduce influenza rates. This is the reason both the CDC and the WHO advised against mask use.

This past winter's flu season was among the worst in recent years and was stopped dead in it's tracks when covid-based social distancing and masks came into effect.
Your own graph shows 2020 flu cases dropping sharply beginning in week 10 -- early March. The CDC didn't change its guidance to advise mask use until early April (the WHO didn't do so until June) and widespread mask availability and use didn't start until the late April - mid May timeframe, depending on area. Mask use could not possibly have caused the reduction.
 
That's not a tangent, it's the heart of the matter. The question is what constitutes an acceptable risk? When compared to all the other risk factors of premature death, Covid is a small incremental increase, and in no way justifies the exhorbitant negative impacts of shutdowns, impacts which have not only destroyed livelihoods, but people's mental and emotional well-being also.

Are there any data to back that opinion up? Covid is now the #3 reason for death (71/100K) behind only heart disease (165/100K, 2017) and cancer (152/100K, 2017). And that with all the current methods of containing it, and in less than a full year of counting. With less social distancing and mask wearing, it would be higher.

Suicide rates are way up, homicides and violent crime are up, depression and other mental disease, divorce, spouse abuse, etc, etc. And don't forget an entire lost year for tens of millions of children, many of which will bear the developmental impact for the rest of their lives.

And what are those numbers per 100K compared to Covid?
 
Status
Not open for further replies.
Back