The app tries to predict blood glucose in the future (drawing on CGM, food intake, lots of parameters and other data) and based on that either automatically injects insulin, or asks the user to confirm.
People need to compile the app themselves because the liability in the case of adverse effects is insanely high.
It’s a highly driven community. I’ve seldom seen such a useful and well supported application of modern tech.
Source: I worked for a company in the diabetes app space.
I know several diabetics with pumps. If that is the type of thing you mean then it seems reasonable. If you mean getting an Amazon health snooping chip to monitor my health "just in case" then the answer would be "no bloody way"
As a Type 1 diabetic I've been watching these hackers with interest. There's clearly some great opportunities for linking continuous glucose monitoring and insulin pumps, particularly for young children.
Compared to old-school closed medical devices it looks like we might get much more data and much more control over what's going on.
But it's easy to see why the commercial companies haven't produced this yet. If a bug in this means you get too much insulin then you'll end up passing out and it could kill you within a really short length of time.
I spoke to one of my doctors about it and he seemed slightly scared at what these people are trying. There have been doctors researching this for years in carefully controlled trials, and now there are hundreds of people just plugging together phones and trying it out for themselves.
I know we're a little ways off, but this seems like a great platform to combine with automatic blood sugar measurement. Imagine taking one pill in the morning that measures your blood sugar and internally administers insulin as necessary. There are already externally attached pumps that do both (my father, a type 1 diabetic, uses one) but having an electronic device permanently attached to your stomach is a bit of a hassle.
I'm a programmer and I've had a type 1 diabetes for the last 21 years. It is one of the most complex things in my life to take care of
when all of the insulin therapies don't really work that well and I don't wake up to the nightly hypos. Luckily I'm in the German insurance system and getting an insulin pump first and later a CGM. Already I'm using an NFC chip in my arm giving results to my phone and from there to InfluxDB and Grafana. Later with a proper CGM I also get automatic alarms which wake me and my partner if the sugar gets to low.
There are great open source systems for us technical people, but the tech is very expensive without an insurance and requires you to solder an extra device to get the values to your phone. At least with an NFC device you can wrap a smartwatch over it to get the readings automatically to your phone, rooted of course.
> Already I'm using an NFC chip in my arm giving results to my phone and from there to InfluxDB and Grafana. Later with a proper CGM I also get automatic alarms which wake me and my partner if the sugar gets to low.
Very cool! I've recently become friends with a type 1 diabetic and have been learning a lot about it myself. It really is an incredibly intrusive thing that involves constant supervision even with automatic pumps, constantly switching spots to place the needle that much stay in him constantly. He's actually getting a service dog in a month that's trained to wake/warn him if his blood sugar levels get too low. Have you open sourced your solution?
What I would like to see is an integrated system where the a continuous monitoring device comes with the insulin pump and communicates with it automatically. I want open data as well and good software, but more importantly, I'd like to have a device that learns from my habits and recommends adjustments automatically. It would save me the hassle of having to pore through my glucose readings and think about everything every 1-2 weeks, which is what's required of me now to maintain my health (and which I'm negligent on).
And I'd like my insurance to cover it, but that's an entirely different story.
Some of those folks have even automated blood sugar detection and insulin management, creating basically an artificial pancreas. It's pretty cool (open source) software if people want to learn more, called Nightwatch.
Modern medical technology relies heavily on computers and software. Take an infusion pump for example. Controlled by a microcontroller and using software. Or insulin pumps; and some vendors are actually considering to add Bluetooth to insulin pumps, so that patients using such a pump can check its status on their smartphone (or on the upcomming smart watches). Also you can adjust the infusion rate of an insulin pump to accommodate for ingested sugar. Overdosing on insulin can send a person into shock and kill.
AI is not necessary for this, at least not compared to the coarse process of guessing energy exertion/food intake/glycemic index of the food in the near future and dosing yourself accordingly, with occasional blood sugar level tests. Further I cannot believe a medical company/standards board would ever approve such a device (with "AI" decisions).
I find it hard to believe that's a real story (although they do use a full name), in North America at least, vendors refuse to automate insulin pumps because of the risk of litigation (too much or too little insulin can cause life changing damage, even be fatal) - and that type of automation would be `if (sugarlevel increasing-over-time) administer-small-dose(fast acting) - resample`. Basically a PID controller (although not exactly since insulin can only reduce blood sugar, so you'd err on the side of caution in your delivery and have big alarms to get human attention if you detect crashing sugar levels). Given legislation around the world is starting to require AI backed processes to explain the decisions it makes, it seems more trouble than it's worth - this is people's lives.
I've had type 1 diabetes since I was 10. With even a semi-automated solution we would greatly prevent the short and long term complications. The biggest issue is that brain runs on glucose. When we can't manage our levels, the brain shuts down. In order to prevent this we grow up paranoid, watching for every sign. Stress, faulty needles etc. all effect us.
Knowing that we can theoretically prevent nerve damage (almost every person with 10+ years diabetes gets it) and other serious conditions (kidney failure, blindness) and practically not being able to do anything is very depressing.
You can't kill yourself writing c++ code, a buggy local server means your webpage is down, you aren't advocating for making your own insulin analyzer firmware, and using it for people whiout diabetes symptoms.
This could be engineered around--mostly the limitations are liability.
If you, the patient, take a glucose sensor, an insulin pump and tie them together with an app, you don't get to sue somebody if something fails and you die.
If a company takes all three of those, packages them together, and sells them, then that company is now liable when one of the pieces fails and you wind up dead.
The sad part is that AI/ML would be incredibly useful for all this. People could take pictures of their meals, upload their sleep and exercise data, and weigh themselves, feed all that into a cloud algorithm that could determine how you are likely to react and dial your insulin plan into your pump while feeding the glucose readings back to the cloud.
Sadly, I suspect none of this is going to happen in the US due to the liability. Maybe China will pull this off.
I use, and have contributed code to, one of these (OpenAPS/oref0). There's actually a pretty shocking amount of spaghetti code, shell scripts, and cobbled together components in different languages. The key thing that makes it work is that, safety-wise, it fails more gracefully than you might think; if it stops working it falls back to what would be the no-automation standard of care anyways, and there are high- and low-glucose audible alarms from the CGM going through an entirely separate path that has no DIY in it.
FDA approval should at least require key algorithms be disclosed in code form.
I’m thinking specifically about “artificial pancreas” aka “looping” insulin pumps, which automatically adjust how much insulin they give you based on a continuous glucose monitor (CGM) providing blood sugar readings.
My kids use these devices, and I’ve watched their devices dose insulin (a potentially lethal drug) erratically when the CGM data itself occasionally goes haywire.
I wish I could review the algorithm for myself. There’s no excuse for a piece of software which is automatically injecting drugs into my children to be a black box.
I have type 2 diabetes and I've always been a bit envious of insulin pumps, not because I love having a machine plugged into my body (I don't), but because they offer better control of blood glucose peaks, since they use fast-acting insulin exclusively. The pancreas secretes insulin in pulses and pumps can mimic that. What has turned me off from pumps (apart from doctors generally not recommending them for type 2 diabetes) is the lack of a sensing mechanism. This DIY system sounds absolutely perfect. It could keep my blood glucose perfect at all times. At the moment, I'm injecting in the mornings and the evenings and have little control over my peaks apart from adhering to a strict diet (and we all know how hard that can be), and every peak does a little bit of damage to my eyes, nerves, etc. I've always wanted to have a more sophisticated treatment option, and this may be it.
Monitoring might be fine, but installing something in the body that can kill you and then running it on node.js doesn't seem like a good idea.
reply