Join 3,552 readers in helping fund MetaFilter (Hide)


Google Gas Permeable
January 16, 2014 4:58 PM   Subscribe

Over the years, many scientists have investigated various body fluids—such as tears—in the hopes of finding an easier way for people to track their glucose levels. [...] We’re now testing a smart contact lens that’s built to measure glucose levels in tears using a tiny wireless chip and miniaturized glucose sensor that are embedded between two layers of soft contact lens material.
posted by (Arsenio) Hall and (Warren) Oates (41 comments total) 9 users marked this as a favorite

 
I would love to have a continuous blood glucose monitoring system like this. Sometimes I like to see how a particular food affects my BG level and that means I have to test at least two additional times per day. My concerns are that BG monitoring capability would make my toric contact lenses even more expensive and that I would need a Google+ account in order to use this device.
posted by Rob Rockets at 5:07 PM on January 16


The next step will be to have microscopic LEDs on the smart contacts, so they can continuously broadcast advertisements important information directly into your pupils 24-7.
posted by happyroach at 5:15 PM on January 16 [6 favorites]


And somewhere deep inside the NSA, they are creating a new office dedicated to analyzing metadata collected from BG-monitoring contacts.
posted by darkstar at 5:27 PM on January 16 [1 favorite]


Oh man, this looks neat and useful and..oh it's developed by Google? Fuck those creepy data collecting jackasses.
posted by Brandon Blatcher at 5:29 PM on January 16 [5 favorites]


If I had to prick my fingers a zillion times a day to test my blood sugar, I would be pretty psyched if I could get better information by wearing a contact lens, given that I would be wearing contact lenses anyway.

On the other hand, if I were a total narcissist who enjoyed quantifying every aspect of my physical existence just because quantifying ME! is cool, I might be pretty psyched, too.
posted by ArbitraryAndCapricious at 5:30 PM on January 16 [1 favorite]


huh, interesting. I've watched abut twenty years of newly-floated non-blood draw insulin monitoring technology get floated and fail over my wife's shoulder. Ten years ago, hearing that Google was headed this way would have made my day. Now, I am actually put off.

It's not all Google's fault. She uses a pump that adjusts based on the traditional finger-prick testing methodology, and the pump gets the data for the adjustments via BT from her meter. But the pump is marketed to be matched with a BT-enabled skin-mount invasive meter, and it sure seemed like she was in the minority in choosing to not go with the continuous monitoring when she went with the pump.

Years later, relying on anecdote rather than data, she made the right choice to stay with a manual-monitoring technology that she mastered as a preteen. Those ancdotes seem to me to indicate that the dual introduction of two new approaches at the same time (continuous glucose monitoring in a home based healthcare setting and wireless data transmission in a home based healthcare setting) may have increased the volatility of the application. She's sure she made the right choice, and I’m happy to support that.

So how does that condition my response to this idea? Well, firstly, I'm just skeptical of non-invasive glucose monitoring now, although I understand why the contact-based approach offers a better sampling environment. Second, I'm suspicious of the reliability of certain local-area wireless connection protocols, now that I've lived with seven years of Bluetooth dropouts. Third, Google has yet to earn back my trust in the wake of G+ and assorted identity management missteps, and I would certainly NOT endorse sharing medical data with the company, probably forever, unless they denounce capitalism and shut down Wall Street or something equally impossible. Those minor management errors fundamentally changed my relationship with the company, and I don't see that changing, basically, ever.

Fourth, indulge me in a joke.

Google Glass, but contacts, which monitors your biochemistry.

You can't possibly be serious.

I know it's not Glass. It's still a Google-developed eyeball-mounted networked information appliance.
posted by mwhybark at 5:38 PM on January 16 [2 favorites]


A friend of ours told us she worries about her mom

The hell? Why does such a neat technological development have to be announced in such a creepily pseudo-folksy way? Google is a giant corporation; it's absolutely bizarre that it should be telling the world about its R&D in a blog post about how its "friend" worries about "her mom."
posted by RogerB at 5:38 PM on January 16 [1 favorite]


I am looking forward to google periscope so I can look down on those who can't afford to spend $3,000 on stupidity.
posted by cjorgensen at 5:42 PM on January 16 [1 favorite]


It needs another component to record what the user was doing at a particular spike or drop or whatever. Without that info, it's difficult to find the cause and effect for BG changes. I

wonder if Google has plans to constantly monitor an individual.
posted by Brandon Blatcher at 5:44 PM on January 16


My wife too has avoided the CGM sensors, mainly because she doesn't want to deal with another infusion set. 19 years of wearing a pump has already caused enough scar tissue in her abdomen area. I always root for these advances, but 22 years of being married to a Type I diabetic has caused me to get very cynical.

I will say I think the next big breakthrough will have to come from outside the diabetes industry. Medtronic is making way too much money on $6000 pumps and all the disposable supplies for which they maintain a monopoly supplier position; and the test strip companies are making way too much on $1 each test strips (my wife uses 8-10 a day).

So go Google.
posted by COD at 5:44 PM on January 16 [5 favorites]


It is cool that you all dislike google so much. But I hate sticking my finger 3-6 times a day and not having real time blood glucose information even more.
posted by munchingzombie at 5:50 PM on January 16 [28 favorites]


In other diabetes news, this article claims an advance on the preliminary-research end of things towards oral insulin.
posted by XMLicious at 5:54 PM on January 16


But I hate sticking my finger 3-6 times a day and not having real time blood glucose information even more.

Nobody likes sticking their finger and real time BG info is only 5-30 seconds away at any given point. I suspect people will put too much emphasis on second by second or minute by minute fluctuations that ultimately aren't important long term.

this article claims an advance on the preliminary-research end of things towards oral insulin.

Yeah, there's usually an article or two a day about the next awesome diabetics discovery that will cure everything, once its out of testing in a few years.
posted by Brandon Blatcher at 6:00 PM on January 16


I suspect people will put too much emphasis on second by second or minute by minute fluctuations that ultimately aren't important long term.

That's why we have smoothing algorithms.
posted by ryoshu at 6:02 PM on January 16


I was kind of a doubter about Minimed's new continuous glucose monitor system, but it's been pretty neat in the few months that I've had it. For those who don't know, there's a small wire filament that goes into the interstitial fluid and measures (via a small electric pulse) glucose levels every five minutes. It's not as precise as blood glucose, but the ability to see trends has been a big step forward. It doesn't make diabetes simpler - it actually makes management a whole lot more complicated, but if you are interested and motivated to handle the additional complexity, with some luck it can be useful.

I'm hopeful that google can do more in this direction, because the lack of a constant, reliable glucose measuring system leads to too many people, adults and children, slipping into a coma or dying from low blood sugar every day.

Hey, COD, I also avoided getting a CGM for years because, like your wife, I already have a lot of problems with the limited real estate on my abdomen and the problems caused by infusion sets. But I have to say, the sensor is really really tiny - like a little thread, and I have not had difficulty with it or with allergic reactions to the adhesive so far, unlike my issues with the infusion sets.
posted by chinston at 6:13 PM on January 16 [1 favorite]


Brandon Blatcher, granted real time BG is only 5 seconds away if you have the time to check it, and some reason to think that you should. But if you're driving, it is not so great to whip out your meter with one hand on the wheel, yet a quick look at a monitor may indicate that you are trending low and so you can correct. Or, on the other hand, I would tend to be drifting up toward 200 after lunch without any noticeable symptoms. Things get kinda crazy at work around 2 or 3 pm, I couldn't easily check all the time, and often these highs might go untreated for a couple of hours. Now with an always-on device, I program it to alarm if it predicts that in 30 minutes my glucose will be heading north of 150. Sure, it would be foolish to focus on second or minute fluctuations, but I think it makes more sense to first develop a technology that would even make that possible, then solve the human cognitive problem of overfocusing on details or something, via smoothing algorithms or other patient education.
posted by chinston at 6:20 PM on January 16 [1 favorite]


On the other hand, if I were a total narcissist who enjoyed quantifying every aspect of my physical existence just because quantifying ME! is cool,

For what it's worth, I recently spent a couple of months doing the quantified life bit, and what I found out was kind of distressing. You might not realize how much of a day you spend basically immobile, for example, and how bad that is for you.

I've since ditched my office chair and moved to a standing desk, made a couple of dietary changes, and routine changes, and I feel way, way better in a lot of important ways. Back problems gone, knee problems gone, sleeping better... Call it narcissism all you want, but there is a lot to be said for actionable data and measurable change.
posted by mhoye at 6:23 PM on January 16 [2 favorites]


I know that when my BG is critically low I completely lack the presence of mind to check my sugar or do anything for that matter. I am dripping sweat, my eyes can't focus, you couldn't get a complete sentence out of me. Just checking my sugar isn't going to alert me that I am about to die. A continuous monitor would.

I think about the dozen times my husband has woken to find me completely unresponsive because my sugar is too low. He has rubbed sugar in my gums to keep me from certain brain damage or death.

Yeah, finger sticks get the job done some of the time. But, you know, let's keep booing because we don't like the company that hosts our email.
posted by munchingzombie at 6:27 PM on January 16 [14 favorites]


Wait a minute this means instead of pricking myself with a little needle I have to STICK SOMETHING IN MY EYE?

THIS IS NOT AN IMPROVEMENT.
posted by localroger at 6:53 PM on January 16


I saw a thing on nova a while back about how they had developed a contact lens that could deliver medicine over time to the eye after surgery.

And another group had developed the ability the put a few digital pixels that could be seen by the wearer. Not clearly but they could see a blur.

I both love and fear this a advancement.
posted by sio42 at 7:04 PM on January 16 [1 favorite]


also, hey, apparently there is a Team D here on MeFi based on this thread. I'm thinking I should at least "contact" y'all if I haven't previously. Can't think of a good in-site way to set up an interest community, but, um, let's think about how that might go down, maybe?

Oh, I know: G+ ;)
posted by mwhybark at 7:11 PM on January 16


Wait a minute this means instead of pricking myself with a little needle I have to STICK SOMETHING IN MY EYE?

THIS IS NOT AN IMPROVEMENT.


This used to be my objection to contact lenses. And five years ago, I finally said "Fuck this noise, specs suck.". I got contacts. The first two days of taking them out and putting them in kinda sucked, yeah, but now I don't even need a mirror to take 'em out or put 'em in, can do it in the dark, on the train, on the bus, and even in the car in spite of my husband's propensity to brake randomly.

A little practice, and you don't even notice any more. Promise.
posted by MissySedai at 7:25 PM on January 16 [1 favorite]


I'm in love with my Minimed/Medtronic continuous glucose monitor, and from what I hear the Dexcom system is much better even than that. Minimed's newest system partially closes the loop toward artificial pancreas by automatically stopping insulin delivery when a low is detected. While I'm extremely skeptical to mildly angry at the FDA push towards fully closed loop systems / artificial pancreas systems, I do think the low blood sugar shutoff would save a lot of lives.

As for the contact lens it sounds lacrimazing, and I'm hoping for a less invasive and/or more accurate cgms sometime.
posted by BrotherCaine at 7:28 PM on January 16


A little practice, and you don't even notice any more. Promise.

Just don't think of that scene from Fire in the Sky while you do it and you'll be fine.
posted by BrotherCaine at 7:29 PM on January 16


While I'm extremely skeptical to mildly angry at the FDA push towards fully closed loop systems / artificial pancreas systems, I do think the low blood sugar shutoff would save a lot of lives.

I'm missing something -- I've wondered why there AREN'T closed loop fully autonomous systems. Seems like the tech has existed to implement an artifical pancreas for at least 10 years. Did a negative get dropped in there somewhere?
posted by localroger at 8:29 PM on January 16


Pretty sure BrotherCaine dropped a negative -- the FDA has been horrendously and notoriously slow to even consider partially closing the loop.

I think this is terrific. I have a Type 1 Diabetic whom I am close to and I know this would make a huge difference in his life. Even if this weren't such a revolutionary device, I'm excited simply by the prospect of an 'outsider' company getting involved. Google in particular seems like it has the resources and out of the ordinary approach to possibly have a chance of getting something new approved in a timely fashion.
posted by telegraph at 8:38 PM on January 16


Why does such a neat technological development have to be announced in such a creepily pseudo-folksy way? Google is a giant corporation; it's absolutely bizarre that it should be telling the world about its R&D in a blog post about how its "friend" worries about "her mom."

It's because people are stupid and increasingly need to be talked to throughout their lives as if they are children.

(and by "people", I don't mean Google, but rather those to whom they are trying to market this).
posted by modernnomad at 8:41 PM on January 16


FDA has been horrendously and notoriously slow to even consider partially closing the loop

Reality check here: the reason why there isn't a closed loop / "artificial pancreas" system anywhere in the world is because continuous glucose monitoring tech isn't remotely reliable enough to seamlessly dose insulin. This isn't a US FDA problem, at least not exclusively.

A few of us in this tread have CGM's attached to us right now and I'd be shocked if anyone would trust theirs to give a bolus of insulin. I have the latest version from Minimed and it truly is good, as BrotherCaine said above, but it is also frequently 10-20% off from my actual BG number. If I took insulin based solely on my sensor I would frequently be unconscious or dead. And the sensors sometimes fail without being obvious enough to throw errors, causing them to be even less accurate. The best tech we have right now will stop giving you insulin after X minutes if a low is detected and not acknowledged - which is great!, but only a first step on a fairly lengthy walk towards a closed loop system. The reason why diabetics and their loved ones are sometimes cynical is because the current system I've described is a 3rd generation sensor; they've been working on this for 10-15 years already.

The newest minimed system a few of us are using was widely released in Europe about 2 years ago. That's the problem with the FDA, in my opinion. On the other hand, there's a pump with integrated CGM from Animas Diabetes (a J&J Company) that was released in Europe 4 years ago and failed FDA testing. The Animas system likely won't be released in the US at all at this point. I have to trust that this is a good thing - someone isn't dead because of this. I think this is really interesting new direction for research, but it may be decades behind the current CGM tech. Insulin dosing requires a high degree of precision, and it turns out that getting to that level is very hard. Sometimes even big extremely experienced companies don't get it right.

In summary - the pace of development with less-invasive glucose monitoring is agonizingly slow everywhere in the world, not just in the US. The best CGM tech requires frequent calibration and is not accurate enough to base medication decisions on. And it sometimes fails unexpectedly for a variety of reasons. There are only about 2 million Type 1 diabetics in the US - not a huge market even with the amount of money we spend on treatment. I hope one way or another we get to a point where more companies enter the market and invest in solving a disease that really is a huge pain in the ass.
posted by pkingdesign at 9:36 PM on January 16 [2 favorites]


This used to be my objection to contact lenses. And five years ago, I finally said "Fuck this noise, specs suck.". I got contacts. The first two days of taking them out and putting them in kinda sucked,

When I got contacts, she gave me the patient education and she's watching me fumble with the thing for a few tries. And she says "Don't worry, it takes most people about 30 minutes to get it the first time." I don't know if this was the plan but I swiftly decided I had better things to do than poke myself in the eye for over half an hour and blasted that bastard in there post haste.

Interestingly, she said usually women are way faster but apparently with that motivator I had not come off so poorly.
posted by save alive nothing that breatheth at 9:46 PM on January 16


Foucault meets Ballard, as dictated to Phillip K. Dick.
posted by Conrad-Casserole at 10:31 PM on January 16


Wow, wait, when did everyone start hating google this much? Like maybe 1/4 of the comments in this thread contain some kind of serious hate.

I noticed that too around the google buying nest news, but yea... When did everyone suddenly turn on them? What news did I miss?
posted by emptythought at 12:29 AM on January 17 [1 favorite]


Why does such a neat technological development have to be announced in such a creepily pseudo-folksy way? Google is a giant corporation; it's absolutely bizarre that it should be telling the world about its R&D in a blog post about how its "friend" worries about "her mom."
Note that it was written by the project founders. I think you underestimate the extent to which, even at a large tech company, projects can be quite personal to their creators and leaders.
posted by Bahro at 12:41 AM on January 17 [3 favorites]


Why does such a neat technological development have to be announced in such a creepily pseudo-folksy way? Google is a giant corporation; it's absolutely bizarre that it should be telling the world about its R&D in a blog post about how its "friend" worries about "her mom."

Because it's a blog post written by someone who isn't a professional writer?
posted by atrazine at 1:50 AM on January 17


Wow, wait, when did everyone start hating google this much? Like maybe 1/4 of the comments in this thread contain some kind of serious hate.


Somewhere around the time they started forcing G+ users to use their real names, not the name they had signed up with.

Which was before Google Buzz was forced on all users as an opt-out feature, which allowed users to see others contacts. Can't remember if that was before or after putting ads in email. T
posted by Brandon Blatcher at 3:53 AM on January 17 [1 favorite]


No, I didn't drop a negative, the FDA has been pushing on device manufacturers in some ways to start laying the groundwork for a closed loop system. In theory that's great, but I think the reality is that in addition to accuracy related issues with CGMS and even fingerstick systems, my big issue is latency of insulin effect plus latency of blood sugar detection. I'm a distracted enough diabetic that I occasionally (once a month or so) fail to deliver insulin before eating, and instead deliver insulin right when the alarm goes off on my pump for an expected high (which given how tight my blood sugar alert thresholds are is probably close to when a closed loop system would actually deliver insulin with today's tech/algorithms). On those occasions when my insulin boluses for meals are reactive rather than proactive (delivered before eating), I tend to go high and stay that way and require a much larger total insulin dose over the course of the day to deal with the same amount of carbohydrates. Taking too much insulin has its own health risks too.

While the FDA has been pushing for a lot of closed loop tech that I suspect won't get used any time soon from device manufacturers, the following things don't appear to be happening: trials of implanted insulin pumps, and integration of Dexcom CGMS into other pumps. Surgically implanted insulin pump trials by Medtronic have basically been suspended in the US, and some of the trial participants in the US fly to France every quarter to get refills because they believe their ability to more tightly control blood sugar and reduce complications is much better with an internal pump. When people are willing to put up with that kind of inconvenience for a perceived gain it makes me sit up and take notice, especially when one of them is a retired physician. Additionally Dexcom's efforts to meet certain FDA criteria for closed loop systems appear to be impacting their ability to integrate their product with insulin pumps now.

I'm not wholly against insulin pump technology, I just think it's risky to dictate that it's going to happen without closing some of the latency between CGMS readings (5-10 minutes behind venous blood readings) and effective insulin delivery with current fast acting insulin (15 minutes to start of effect, and 45 minutes to peak onset). Admittedly, if Viaject insulin (or whatever they are calling it now) reaches approval that may close the gap enough to make it worthwhile, as well Afrezza could make a closed loop system worthwhile, although I'm not sure how you incorporate inhaled insulin into a closed loop system. Without a decrease in latency, I think a fully closed loop system (no manual bolus) is going to result in more sticky highs, more complications, and issues with blood sugar volatility. I'm just going from my own experience with trying to control T1 with a CGMS and pump. It may work out a lot better for your average diabetic.

While I've talked to reps from Dexcom, Animas, Medtronic, Roche, T-Slim, Omnipod, and know a little about the Asante pump, I don't have enough insider knowledge to know for sure that the FDA pressure is focusing pump development the way I suspect. Certainly Medtronic had other reasons to drop internal pump efforts (surgery is a hard sell), and it's not like Dexcom was integrated with other pumps even before the closed loop efforts really started to heat up. Some of the information I use to analyze these things is anecdotal or dated.
posted by BrotherCaine at 4:50 AM on January 17 [2 favorites]


TLDR; I worry that the FDA is losing focus on safety and delaying time to market to pursue a closed loop system that may never work.
posted by BrotherCaine at 4:57 AM on January 17


I meant to say I'm not wholly against closed loop pump tech.
posted by BrotherCaine at 5:41 AM on January 17


Thank you, BrotherCaine, that's a useful perspective. The problem that I see with a complete closed loop system, and which I don't really see how you overcome via software/programming, is the following situation, which in my case anyway is quite common - at time T-0 you eat a medium or high carb meal, and then within 1-3 hours, engage in vigorous exercise. As long as the exercise takes place while a significant amount of insulin is still "on board," then in my case anyway the result would likely be a severe low. I don't think I'm alone in this. I guess otherwise as long as there's no exogenous effect on the system, then a closed loop set up would in theory work. But unless insulin gets a lot more rapid-acting, or there is some system of introducing glucagon automatically, no closed system is going to be able to do without human override and management.
posted by chinston at 6:24 AM on January 17


Why does such a neat technological development have to be announced in such a creepily pseudo-folksy way? Google is a giant corporation; it's absolutely bizarre that it should be telling the world about its R&D in a blog post about how its "friend" worries about "her mom."

I am very skeptical about Google, but their choice here resonates with me. A relative has diabetes and does the finger-prick thing every day, plus shots before meals. Sometimes he forgets, so we ALL have to try to remember to help him remember. At every family dinner there will be a couple of rounds of "Did you do your shot?" (and the second one is usually answered by a testy chorus of voices).

Second, as I understand it, there's a circular problem that when you forget to do a test, that's when you need to remember to test -- but that's also when you just forgot, and are less and less able to remember. When $RELATIVE gets a little cranky between meals, someone will eventually ask, "How's your sugar?"

So I would love for us all to stop having to remind him, and I would also love for him to get a more consistent level of sugar in his body. The fact that he loves technology would make something like this (though maybe not actually contact lenses) great for him.
posted by wenestvedt at 7:42 AM on January 17


pkingdesign,

The fingerprint systems are only accurate to about 15% (source: someone who spend a decade building them).

Go play with strips from different lots, maybe from different manufacturers.
posted by effugas at 11:19 PM on January 17


Google Glass just got a lot less geeky: The headset of the future now works with prescription lenses
posted by homunculus at 1:58 PM on January 28 [1 favorite]


« Older Elena Shumilova photographs her sons and their ani...  |  Craigslist Mirrors (SLTumblr)... Newer »


This thread has been archived and is closed to new comments