I find some of the this vague in this announcement. Non invasive BGM is the "holy grail" in this field. Tons of companies have failed after big investments. Some scammers are out there they claim since decades they solved the problem. I would be curious about peer reviewed publications of their technology.
A friend of mine is working on this too. Close to market. He is the only one that regularly published peer reviewed publications on this topic.
In fact, I remember I have written an SBIR grant proposal on the technology that was evaluated by an idiot who googled one of the components, realized a single unit (as compared to 10000 units) would cost 50 grant and rejected it based on the claim that the technology would be too expensive in practice. I had written in the proposal that, produced in quantities, the cost of this item is in the mid 3 dollar digits per unit. An idiot and google are a very dangerous combination, even at the NIH.
When I started my Phd in 2011, one professor, an absolute top player in the field and prof in a top tier university, revealed that he had patented a fully functional prototype a few years back, which allowed similar cheap non-stick measurement of glucose (i think this was on a principle similar to pulse oximetry rather than RF).
The patent was swiftly bought by a large pharmaceutical company and "buried", in order to protect sales of blood-sticks; by far the main source of profit in diabetic devices (his words).
Doubt it. There are so many issues with any technology regarding non-invasive. Is it still accurate if the individual has a fever? etc. etc. There were at least 50 companies that tried to build such devices, some start-ups had funding in the 50-100 million USD range - and failed. Samsung abandoned it. Apple is rumored to work on it.
Also, without clinical trials you really can't claim this device is reliable. You would also have to compare to real blood glucose and not the finger prick test.
Forgive my ignorance but, isn't a finger prick measuring real blood glucose? Or are you referring to a more long term/average a1c? (Or is there something in between that I just don't know about?)
"Real" blood glucose is glucose as measured by a reference standard device, like the YSI 2500, arterial blood gas machine etc. Finger pricks are accurate to within ~15% of the real blood sugar, and reference machines like the YSI are accurate to within ~1%.
Exactly. And I doubt that this device is used in any lab, except in a clinical lab setting. I have seen it in clinical trials. Focus on the word "arterial blood", this gives you an idea. :-)
Sure, it's a very bulky device and expensive to run! But, when you're talking about 'real' blood glucose, it's the gold standard that you need to use to pass regulatory scrutiny.
Sounds like you didn't idiot-proof your proposal. A lot of good paper writing guidelines, like explain value of contribution fast and early, could be phrased as idiot-proofing.
Maybe. But maybe you don't know how SBIR grants work. I consider them highly corrupt and I was told by people who got them how they got them and they even advised me, not to waste my time (two person, both hat P1 and P2).
The solicitation was for non-invasive BGM. Of the accepted proposals, nothing came out. Big surprise. At least I got a very high rating for "originality" of our approach.
I think they mean the NIH grant reviewer dismissed the grant request on the basis that the technology was too expensive - ignoring that the research was intended to catalyze mass production which would have made the technology economical. Thus missing the point of the grant proposal.
Yes. Basically. This item is available for a lab setting with many fancy options. Cost: 50.000 USD
If 10.000 units of a customized simple version are ordered, the cost is 500 USD, give or take a few hundred dollars. This was not my estimate, this was the quote given by a company.
Hi, this is actually my field of expertise! IdealMedTech is a clinical glucose control startup targeting inpatient glucose control.
Having an MARD (mean absolute relative difference, or in math terms, E[|g-r|/r], g is measured glucose, r is reference glucose) of below 15% is actually considered very good for glucose monitoring.
We did a study that involved modelling different levels of sensor error, and you actually start to see diminishing returns on control once your MARD passes below 10%, with 15% being a good target.
Most sensors on the market that are considered best in class (eg Dexcom G6) have guaranteed MARD below 15%, usually sitting right at 10%. Getting lower than that is very difficult for interstitial glucose, especially using the classical glucose oxidase sensing mechanism.
There are other ways to sense glucose using, for example, Raman spectroscopy, and there have been companies successful in employing these techniques (OptiScan comes to mind), but I'm not yet convinced by the capacity of spectroscopy to deliver accurate results in the presence of interference from drugs and perfusion issues.
> Most sensors on the market that are considered best in class (eg Dexcom G6) have guaranteed MARD
Disclaimer: I work for Dexcom.
A few times every year, there's a ton of hype about some new noninvasive device, and the newbies get a bit worried, but the old timers just shrug it off.
We (humanity) will literally be able to regrow new pancreases with stem cells before we have accurate noninvasive blood glucose devices. Even Tim Cook with unlimited resources at his disposal admitted it's a much harder than they thought.
That said, I think there's a large market for people without diabetes who are interested in their blood glucose and who may not need super-accurate readings. For example, as a long distance runner, it'd be super helpful to know if my blood sugar is trending down before I bottom out, so I can know when to eat a gel pack, etc. It would also be useful for people trying to lose weight to learn how their eating habits and exercise affect their blood sugar levels.
Noninvasive devices may be better for this group of people who don't need the devices to be accurate enough to make decisions about insulin-dosing.
"A few times every year, there's a ton of hype about some new noninvasive device, and the newbies get a bit worried, but the old timers just shrug it off"
This.
But a buddy of mine may surprise you soon. And I worked in the field too.
And a lot of people are waiting to be (surprised). It super duper helpful for t2dm folks to know the "trend" - which food works better, which food doesn't, is the medication working, how much workout is required to drop the blood sugar levels, the dawn phenomenon etc.. with literally billions of people getting affected with t2d it would be a game changer. Still hoping for apple watch xx to have that feature.
Thank you for your work. My 9 year old son has worn a Dexcom for the past 3 years (Medtronic prior to that) and it's a godsend to be able to pair it with Sugarmate.io so we get phone calls in the middle of the night if his BG goes low.
In addition, I imagine confounds which are patient/location specific can be calibrated out to some extent, by comparing against finger-stick while fasting, and longer-term A1C.
The whitepaper gives actual prep methods for tissue phantoms. It shows a clean correlation of ca 1.72GHz transmission to glucose. Presumably altered parameters in tissue phantoms could help roughly characterize confounds.
This... brought back some memory...to average or not to average... C8 Medisensors achieved better weekly MARD in the early 201*. But the problem was, on a bad day, no one knows why it was bad..., so no way to alert the user (unless the reading was way off). Even though on "average", performance looks nice, but a bad day the data can be very problematic for a user, and the user has not way to trust the device based on MARD...
This is a great point, and precisely why we're using two sensors and a simple voting algorithm in our first in man trial (3+ sensors and associated algorithms were deemed too risky by FDA for initial studies).
It's not the number of sensors, it's the novelty of the voting algorithm required to fuse those sensors that gives them pause. FDA likes to be _VERY_ conservative when it comes to trying new things. We're trialing our control software, which is already novel to them, so anything we can do that's not brand new helps reduce our risk.
More information should help make more informed devision, if the algorithm can not clearly explain it to FDA or FDA cannot understand that is the case... something seems wrong here... all the best wishes and also encourage more exploration.
Patients averse to finger sticks are going to use this much more often. It's going to provide more useful data to them. Much of the time the exact number isn't as important as knowing the trend or the general range.
Whether this makes more sense than a Libre for someone just depends on price and how insurance treats it.
I tested my blood sugar when my wife was pregnant and she was testing hers. I was amazed at how just from the small blood stick, and even if I had my wife do it and didn’t see any blood, I’d get sick to my stomach every time I did a blood stick. I’m glad I don’t have diabetes or I’d have a terrible time of it.
Yes, it's not fun. On top of that, for a permanent diabetic, finger sticks over a long period of time are an infection risk, especially if the patient isn't good at using all of their fingers (I got comfortable with it and there were still at 4-5 finger locations that I dreaded pricking.)
Agreed, as a diabetic and climber I've used a lot of the libre pucks but honestly, there is no spot on my body I wont use while practicing my sport. Still the libre was miles better than always pricking my fingers and arms.
I've been thinking about what kind of EM-signal could be used for this for a long time, IR-spectroscopy never seemed possible. But this has real potential.
I worked on the Freestyle Libre, I’m thrilled to hear it’s at least a workable solution for you. When I was working at Abbott, I often volunteered for testing finger-stick devices, which made me appreciate a small part of what people with diabetes have to endure.
I am deeply grateful for your work and that of Abbott. The Libre has been life changing for me, possibly ever more so than getting a pump as it gave me a much clearer picture how my body reacts to different foods, exercise, lack of sleep, etc.
After about two years of usage I have encountered very few issues at all. My only complaint was that I would like to scold whoever crippled the sensors and readers somewhat by making them region specific for what I can only assume are purely market control reasons (not to mention that this is not obvious from the packaging or instructions from what I can tell and I only found out when talking to an Abbott representative that cautioned me that readers and sensors were not compatible between regions).
I use a Freestyle Libre sensor and I like it better than finger sticks. When I was doing finger sticks I was averaging 4 tests a day; before each meal and at bedtime. With the Freestyle Libre I'm checking my glucose an average of 20 times a day. But the Freestyle Libre sensors are quite expensive. I have insurance so my out of pocket expenses are very low, but my provider bills $1074 for 2 sensors, which last for 4 weeks.
I have read comments from other posts that buyers without insurance can get sensors for substantially less, but they are still quite expensive. When I was doing finger sticks I could get enough test strips for a month for about $30.
“We know that not all people with diabetes are looking for a wearable continuous glucose monitoring device to manage their diabetes."
Okay that's pretty bogus. Some people can't afford it, but everyone with diabetes would benefit from a CGM. There's much more to be gained from trends in CGM data than singular data points. (I was diagnosed with Type 1 at 16, been on a CGM for most of that time).
The Dexcom G6 is the best CGM on the market, but its sensor filtering (Kalman or particle?) has smoothness and/or continuity priors that cause persistent inaccuracies.
I observe these persistent inaccurate states on the fifth or sixth day using a sensor. Calibration doesn't help. The reading goes wrong and stays wrong.
It probably happens to you, too, but anyone who doesn't test with a glucometer will likely not notice the failure. You will be harmed by these inaccuracies. Persistent inaccuracy is different than the Gaussian (normal) error of a glucometer, which is far more benign.
Also, CGMs are slow. The reading is based on interstitial fluid and lags direct blood glucose test by 15 minutes or more.
I have completely abandoned the Dexcom G6 and have gone back to 16x daily finger stick blood testing. I am also on a low-carb/high-fat diet, which makes it easy to maintain normal human blood glucose. My HbA1c has been 4.9 for a decade.
I was overjoyed to use the CGM at first, but it turned out to be a net loss. If you want to control your diabetes, use a glucometer and adhere to a strict low-carb/high-fat (ketogenic) diet.
I've done accuracy studies of the G6 for a submission to the FDA, and while it's true they do filter out spikes, they're never more than 5mg/dL off from the measured value (which itself is usually within 10% of the real value). Interstitial lags are closer to 5 minutes when you're not experiencing severe perfusion issues, which is more than close enough for making dosing decision, especially considering subcutaneous insulin has similar or longer lag times (depending on lots of things of course, and subject to an individual's own body).
Any diabetic can confirm for themselves my claim about the Dexcom G6's unacceptable persistent error states.
Simply compare the CGM with your glucometer (testing 8x per day, for example), and watch the CGM go wrong (and stay wrong) on day 5 or 6 of using the sensor.
I wish what you said was true. I would love to have a CGM that works better than a glucometer, but they don't exist yet. In particular, the Dexcom G6 is inaccurate in a harmful way.
I’m T1 and I use the G6. I used to regularly check against a glucometer multiple times a day (without calibrating) to check accuracy, and found no issues. A blanket statement such as “the Dexcom G6 is inaccurate in a harmful way” is certainly not true in general, even if true for you.
Is there a good source for info on ketogenic diets for type 1 diabetics? My youngest has been on pump therapy for the past 5 years and is completely burnt out from all of the alerts and scares (some of it is self inflicted). She was on a medtronic closed loop system for 4 years and that was the source of most of her frustration. She's on g6+tslim now which has been much better, but she still falls off the wagon regularly.
I just think a keto diet would make it easier for her, but I'm seeing all sorts of conflicting info on it and her doctors don't support it. I've done it a few times in the past and getting started is the hardest part. I'd like for her to try but I'm getting no support on the idea. ?
Anecdotal, but a Keto diet has completely flatlined my BG — in a very good way. I’m talking about being in the range of 4.8 to 5.8 for weeks on end. My endocrinologist thinks I’m a wizard and has never seen anyone with 100% in range.
I also love that I never have to bolus with food. It’s the closest I can get to pretending I’m not a T1D. I just eat. If I’m full, I stop and don’t have to worry that I accidentally took too much insulin. To reiterate, I take zero boluses when I’m on the keto diet.
Doctors are against it because they’re dealing with ketoacidosis on a regular basis. Ketosis also means your body is producing ketones (which I believe is the source of confusion for doctors), but that’s fine because your BG is in a normal range. Technically you’re at higher risk for ketoacidosis when on the keto diet because of those ketones, but realistically your BG is not going to reach >15mmol/L so you don’t have to worry too much.
By the way, you may be interested in Loop[0] if you wanted to look at a more advanced closed loop system. It can handle both auto basal and auto bolusing.
> To reiterate, I take zero boluses when I’m on the keto diet.
I'm very surprised by this and would love to hear what you eat while doing keto. I'm T1 but still have to bolus when doing keto, even when doing zero carb and fasting for 18+ hours a day. I only need between 1 and 3 units but there's rarely an occasion when I don't need to bolus at all.
> Doctors are against it because they’re dealing with ketoacidosis on a regular basis.
The issue my endo has is that my BG is much lower than he would like. I've heard the phrase "too well controlled" so many times and it irritates me no end. His concern is that I'll lose awareness of low BG, and can't comprehend how I'm able to function as usual even when my BG is ~2.5 mmol/l (to clarify - I don't aim to run that low; my target range is 4-6).
Could you please share a bit more information about the insulin (kind, amount) you use during the keto diet? and how did you discover the right amount of insulin?
It seems to me, if one would let the insulin slip too low, one would be in danger of ketoacidosis, isn't that so?
I'm on a keto diet for chronic insulin resistance and obesity, caused by 50 years of high carb eating. (Every meal a sandwich, or including potatoes or rice, and lots of sugar.)
I got a CGM a few months ago to help me directly see the impact of the foods I'm eating. It has allowed me to fine tune my diet and get rid of everything that causes a noticable rise in my glucose. My daily graph now shows little more than than the natural daily cycle of glucose produced by my liver: starts rising before dawn, peaks around 10am, drops to my average by 6pm, and drops to daily low around 3am.
If I were a type 2 diabetic (I was headed there), I wouldn't need supplemental insulin because I'm not eating anything that increases my glucose. The only glucose in my blood is the small amount my liver puts there because I need it.
If I were type 1, I'd need to take insulin because my body wouldn't be making any at all. With my natural, consistent, and small-range daily pattern, it would be easy to predict my needs and take the right amount at the right time.
Everybody's body responds differently, so you have to experiment to see what works for you. The CGM has been a great tool for that. I also use a Keto-Mojo for blood glucose and ketone testing now and then.
What's worked for me is a diet of eggs, cheese, bacon, chicken, ground beef, ground pork, roast beef, turkey, pepperoni, kale, cabbage, Brussels sprouts, coffee, seltzer, bone broth, half & half, heavy cream, rebel ice cream (but not other keto ice cream brands), almond flour, ground pork rinds (as a bread crumb substitute), and probably some more I'm forgetting.
Typical advice is to stay under 20g net carbs, but I've found that I do better if I count total carbs instead for any non-veggie foods. Keto-friendly substitutes for flour and sugar mess me up most of the time, but I've found a few things that are ok. I also do intermittent fasting, only eating one or two meals during a 6-8 hour window, but I do much better if I fast for 36-48 hours. Dr Eckberg has some good videos on why fasting that long makes a difference.
Wow - thought I'd never find anyone else with my diet on HN, but (other than half & half, ice cream and Brussels sprouts), we're surprisingly close. I'd also add a few things, such as almond milk, homemade mayo, flax, cocoa powder, 100% cocoa squares, and head lettuce for burger wraps, but that's about it. Like you, I tried things like carbalose flour and erythritol with bad results, but stevia powder seems ok. I'm glucose intolerant but thin and would likely need some sort of drug if I wasn't eating this way - which I have been since 2008.
I did something similar for a while. Loved it. Great results.
How's your cholesterol, risk for heart disease, etc?
Having high LDL, I switched to a high fiber diet. My LDL hasn't budge.
So now I'm pondering what's next.
I have almost zero knowledge, opinions, ideas about what's best, for me or anyone else. The stuff I've read about cholesterol just confuses me.
One of the few things I did learn (about myself) is that calorie restriction works great. After a few days I barely notice. But adherence is very difficult when I'm in pain.
T1 here. I use OpenAPS with a G6 for closed looping (it's substantially better than Medtronic's looping algorithm), combined with a keto diet. The keto diet makes things a lot easier; there's no carb entry to do, so no opportunities for mistakes. I would indeed recommend it. Just make sure to count calories (at least initially) and keep them high enough; most people on keto diets are using them for weight loss, and that style of keto diet isn't sustainable.
Well, once you've finished "eating" your body fat, you can't sustain the calorie restriction. But at that point there's no reason too. For those of us prone to insulin resistance, we have to continue to eat low-carb to avoid putting the weight back on. (Very similar to how an alcoholic who's gotten sober has to continue to avoid alcohol.) But we don't need to have a caloric deficit anymore.
Your fat sources should be nuts and olive oil and chia seed, with some cheese and egg.
Healthy food with lots of healthy fat.
The brain uses ketones as fuel, so you will find that hypoglycemia is easily handled. In ketosis, your brain will have plenty of fuel, even if your blood glucose drops too low.
I can't say enough good things about my G6. It literally changed my life as far as how I interact with my diabetes.
- I feel safer when I go to bed, because it's there to wake me up if my blood gets low (or too high).
- I feel safer in general because I don't need to worry that my blood is low but I haven't noticed it; and that I'm going to just fall down face first while walking across the room.
- Testing my blood is just a matter of looking, vs going over to (or downstairs to) where the tester is.
- (Because of the above) I have a better "feel" for how my blood sugar reacts to various foods and activities.
- When I'm having trouble with my sugars (morning highs, for example), I can see what my history has been via nice graphs on their web site.
Honestly, it's amazing. While I respect your choices, I can only assume you weren't taking full advantage of just how wonderful having the G6 is.
My numbers tend to range in the high 5s to low 6s. That being said, the past two years have been rough because of a thyroid issue, plus 2 different drugs that impact blood sugar a lot (my blood sugar was hovering around 180 all day today, having not eaten since last night, and taking 80 units of short acting insulin over 4 shots... its been pretty crazy)
I think the usual recommendation is before meals, two hours after meals and before going to bed. So that was 7 times a day for me. The big problem with testing more times is that you are literally damaging your fingers and hoping that they heal before you need to use that area on your finger again. For me this didn't always happen and is one of the reasons I love my FreeStyle Libre.
I just started testing glucose this last week, for health and training purposes. Of course, it just started to get cold in the mornings, and I wake early, so I’m dehydrated, cold and stiff. It takes a bit of water drinking and bodyweight-squatting and arm rubbing to get vigorous circulation for a good drop of blood. I haven’t figured out how to store and dispose of the lances yet, and the bother of it all makes me very sympathetic to those who aren’t doing it for fun and gains.
For disposing of lances - if you're in San Francisco at least, Walgreens provide a free sharps disposal service. They provide a sharps container at no cost, and will take in your full (used) sharps containers.
"...the radiofrequency sensor technology was able to measure glucose levels with accuracy comparable to that of Abbott’s Freestyle Libre continuous glucose monitor..."
Then this means it won't eliminate fingersticks 100%. I've never used the Libre, but I use Medtronic's Guardian sensor, which requires calibration twice a day via fingersticks.
CVS sells their strips at $0.20 once you buy 200+. I'm sure you can get lower.
And why shouldn't this compete ? Their costs will be lower than test strips, their profits could be very significant.
They probably monetize with the device itself and then a monthly subscription for their tracking portal. Sounds quite reasonable to me. As a patient, I'd love it.
Edit: As a startup, it's not your job to match existing offerings - it's to disrupt. Ask Space-X, they can't compete with Boeing either ;)
A friend of mine is working on this too. Close to market. He is the only one that regularly published peer reviewed publications on this topic.
In fact, I remember I have written an SBIR grant proposal on the technology that was evaluated by an idiot who googled one of the components, realized a single unit (as compared to 10000 units) would cost 50 grant and rejected it based on the claim that the technology would be too expensive in practice. I had written in the proposal that, produced in quantities, the cost of this item is in the mid 3 dollar digits per unit. An idiot and google are a very dangerous combination, even at the NIH.