Apple Watch can detect diabetes with 85% accuracy

[bcredeur97]

85% isn't quite good enough for a medical problem IMO but good that progress is being made towards that.

[Doobeedoo]

That's neat, I'm not a diabetic but I do have low blood sugar which I need to be wary about. Not skip meals. 

[Commodus]

10 hours ago, Ryan_Vickers said:

That may be but the fact remains it looks like a very chunky lump strapped to your wrist - almost any other smart watch looks better and more professional imo

The problem, as I see it, is that most of those other smartwatch brands make an incredible number of compromises just to say their device looks like a conventional watch.  "Look at how much it resembles your old watch!"  Yeah, but the battery life is typically lousy, it only fits a small amount of information on screen and it's half an inch thick.

 

I won't deny that the Apple Watch looks like a piece of tech on your wrist, but it can look nice with the right case/band combo.  And importantly, they're actually size-appropriate for women and anyone with thin wrists.  It's still shameful that most Android Wear/Tizen watches are these massive beasts clearly designed by men, for men... and clearly men that never have to tuck a watch under a dress shirt.  More importantly, the Apple Watch actually works well as a smartwatch.  Good battery life (on Series 3, at least), lots of info on screen, a healthy app ecosystem, plenty of fitness options, you get the idea.

 

On the original subject: I'd say smartwatches are best at detecting signs of short-term trouble, like heart arrhythmia and even the prelude to heart attacks.  There have been numerous instances of smartwatches saving people's lives.  While 85 percent accuracy for diabetes diagnosis is promising, that's not reliable enough for definitive action.  Hopefully there can be improvements in hardware and software to boost that rate.

[vanished]

2 hours ago, bcredeur97 said:

85% isn't quite good enough for a medical problem IMO but good that progress is being made towards that.

 

2 minutes ago, Commodus said:

[...]

 

On the original subject: I'd say smartwatches are best at detecting signs of short-term trouble, like heart arrhythmia and even the prelude to heart attacks.  There have been numerous instances of smartwatches saving people's lives.  While 85 percent accuracy for diabetes diagnosis is promising, that's not reliable enough for definitive action.  Hopefully there can be improvements in hardware and software to boost that rate.

 

And on that topic, apparently it's actually much worse than that.  The 85% is the number of correct diagnoses (telling you that you have diabetes and in fact, you do), but that doesn't factor in the other possibilities.  This post here explained it, but the TL;DR is the number you are probably thinking this is talking about (and rightly so) is actually more like 37%.

 

Spoiler

19 hours ago, Energycore said:

Reminds me of a video from Veritasium on why you have to double check stuff like this: 85% accuracy on the Apple Watch test is a lot less than you'd think.

 

So just because the watch says you have it, doesn't mean you have it. Go see a doctor if it does.

 

UDPATE: Here's the actual probability that you have diabetes if you're diagnosed by Apple Watch, and live in the US

P(D): Probability that you have Diabetes (living in the US) = 9.4%

P(D*): Prob. that you don't have Diabetes = 90.6%

P(W|D): Prob. that Apple Watch says you have Diabetes, if you do have it = 85%

P(W|D*): Prob. that Apple Watch says you have Diabetes, but you don't = 15%

 

The computation then goes

 

P(D|W)=[P(W|D)*P(D)]/[P(W|D)*P(D)+P(W|D*)*P(D*)]

 

Which computes to

 

P(D|W)=[0.85*0.094]/[0.85*0.094+0.15*0.906]=0.3702...

 

So, assuming these calculations are correct, and Siri on your Apple Watch goes and says "hey, be careful there's a chance that you have diabetes", then you'll actually have diabetes 37.02% the time.

 

On the other hard, if you go see a doctor and they conduct tests that end up positive, then you're most likely to have it.

 

So please folks, double check every time. It's worth it for your health.

 

EDIT: Had some numbers wrong

 

[Vantage9]

22 hours ago, DrMacintosh said:

This really does show that the Apple Watch was always geared towards fitness over being a “smart watch.” 

 

No, it doesn't.  It shows that people are able to take existing sensor technology and do amazing things with it.  Its the same sensors in all smart watches basically; Apple did not 'plan' or 'gear ahead' for this. 

 

You've already gotten owned by other people showing that this study applies to all smart watches, but I thought I'd jump in as well.

 

 

[Eduard the weeb]

22 hours ago, DrMacintosh said:

This tech can can be added to Android Wear (and presumably Gear watches) as well but in my opinion their market presence makes support for those watches mostly being an after thought. 

smartwatches in off it self seem like and after thought in tech in general imo

[DrMacintosh]

1 hour ago, Vantage9 said:

You've already gotten owned by other people showing that this study applies to all smart watches, but I thought I'd jump in as well.

Don’t think I can get “owned” if I said that in my original post...... rather it shows people didn’t read. 

[Commodus]

10 hours ago, Eduard the weeb said:

smartwatches in off it self seem like and after thought in tech in general imo

Oh, the smartwatch market is doing well -- for Apple.  It shipped about 50 percent more watches during the holidays than it did a year ago.  It's most everyone else who's floundering.  Google still has no idea how to foster Android on anything that isn't a phone, and Samsung has an update cycle that could best be described as... relaxed.

 

The smartwatch market right now is more like MP3 players when the iPod hit the mainstream -- it's not even so much that Apple is making the best possible product as that its rivals just don't know how to mount an effective response.

[mr moose]

15 hours ago, LAwLz said:

I was at a seminar about AI where they talked about this and chances are, you're already better off following the advice of a machine rather than a doctor.

In fact, a lot of times doctors are already relying on machines.

 

A doctor only has their experience to go on, and a lot of times that's guesses based on indicate information.

A computer could scan through hundreds of thousands of people with similar test results as you to get a much more accurate prediction.

 

Not saying that you shouldn't trust doctors, but we are already at the point where doctors augmented by machines are better than just doctors, and in a few years removing the doctor completely could increase the number of successful diagnoses even further. Kind of like how future cars will be safer than human driven cars.

 

But with all that said, I wouldn't trust an Apple watch for diagnosing me, even a few generations from now. The machines doctors use are quite different from smartwatches still.

 

So long as the machines are heavily scrutinized for best analytical practice and doctors don't become lax thinking the machine will work it out. 

 

[mr.squishy]

Most Doctors can do this too.

[suits]

My eye doctor can detect it with near 100% accuracy. Not sure which is more impressive. Lol.

[Noctus]

15 hours ago, DrMacintosh said:

Don’t think I can get “owned” if I said that in my original post...... rather it shows people didn’t read. 

Funny, when i read your OP it never originally had it in there. Just accept you got caught.

[Vantage9]

On 2/8/2018 at 3:00 PM, DrMacintosh said:

Don’t think I can get “owned” if I said that in my original post...... rather it shows people didn’t read. 

Yeah, okay.  That's why nobody saw it at first, and that's why your subject line is SUPER DUPER misleading.  

 

This is some apple fanboy crap.  We all get it.

[Jito463]

On 2/7/2018 at 3:25 PM, Master Disaster said:

I always find it incredible how many people are willing to blindly follow the advice of a machine without consulting any kind of expert first.

Yes, rise up with me and resist our coming AI overlords!  Throw off the shackles of virtual tyranny!

On 2/8/2018 at 7:01 AM, LAwLz said:

Kind of like how future cars will be safer than human driven cars.

 

On 2/8/2018 at 10:19 AM, Commodus said:

It's still shameful that most Android Wear/Tizen watches are these massive beasts

When I first glanced at this, I misread the last word.......needless to say, it didn't make much sense.  Especially when the line underneath it mentioned dress shirts.  I was rather confused for a moment.

[LAwLz]

1 hour ago, Jito463 said:

Are you honestly saying that you do not believe self driving cars will ever be better drivers than humans?

[Jito463]

46 minutes ago, LAwLz said:

Are you honestly saying that you do not believe self driving cars will ever be better drivers than humans?

Depends on the driver.  Having said that, with the report showing the massive number of times human drivers have needed to intervene and take over?  I'd say that no, they're not ready for prime time (and won't be anytime soon).

[79wjd]

7 hours ago, Vantage9 said:

Yeah, okay.  That's why nobody saw it at first, and that's why your subject line is SUPER DUPER misleading.  

 

This is some apple fanboy crap.  We all get it.

In fairness, it could very well have been there and 99% of people would have missed it as they read the title and then skim the post at best and skip over the post at worst.

3 hours ago, Jito463 said:

Depends on the driver.  Having said that, with the report showing the massive number of times human drivers have needed to intervene and take over?  I'd say that no, they're not ready for prime time (and won't be anytime soon).

I don't think it's ready YET, but at the end of the day I think self-driving cars will be superior as a whole -- on an individual by individual basis you might find that a very small percentage of drivers are actually better, but certainly not a large percentage (again, once it's working well). There also needs to be more heavy integration -- like smart roads, and of course self driving cars will always be at a disadvantage when the roads are still unpredictable morons.

[ARikozuM]

But can it tell if I'm choking?

[mr moose]

32 minutes ago, djdwosk97 said:

In fairness, it could very well have been there and 99% of people would have missed it as they read the title and then skim the post at best and skip over the post at worst

.

Aren't all edits visible to the mods?  can't they just tell us if it was edited?

[DrMacintosh]

Why is this thread still active

[mr moose]

I just read the OP again,  even now it doesn't acknowledge the study included android wear and the percentage split is unknown. In fact there is only one line concerning android and It insinuates quiet the opposite,  that there isn't anything on android but that the tech can be used buy android wear.  I guess the fact it already is used by android wear and was part of the study escaped the OP.   So regardless if it was edited or not, the OP still is very biased and is not reflective of the actual news. 

 

 

Edited February 13, 2018 by GoodBytes

[GoodBytes]

Just now, mr moose said:

I just read the OP again,  even now it doesn't acknowledge the study included android wear and the percentage split is unknown. In fact there is only one line concerning android and It insinuates quiet the opposite,  that there isn't anything on android but that the tech can be used buy android wear.  I guess the fact it already is used by android wear and was part of the study escaped the OP.   So regardless if it was edited or not, the OP still is very biased and is not reflective of the actual news. 

 

 

Fitbit is very popular, and is part of the study. It is low cost, actually a sport smart watch, and has an app with a focus on sports tracking. Well reviewed, and low cost (compared to many smartwatches).

[GoodBytes]

16 minutes ago, mr moose said:

Aren't all edits visible to the mods?  can't they just tell us if it was edited?

Based on the logs that I am seeing, DrMacintosh did not edit the original posting or title.

[mr moose]

2 minutes ago, GoodBytes said:

Based on the logs that I am seeing, DrMacintosh did not edit the original posting or title.

That settles that then.  Now people can get on with separating the facts from the fanfare.

[MeDownYou]

How useful is this really though? Any doctor nurse can probably get an 85% accuracy rating by looking at you and asking what you had for lunch.