Maybe an artificial pancreas IS possible!
I've had type 1 diabetes for a long time and I've spent much of that time hoping for, and thinking about, a possible cure for diabetes. But I'm also very interested in the research that JDRF is doing on an artificial pancreas.I can understand how a closed loop between a continuous glucose monitoring system and an insulin pump could work to control my basal (background insulin) rate. But I really didn't see how it might also be possible to effectively give boluses (larger doses for meals or BG correction) with such a system.
Yesterday I stumbled across a short paper in Diabetes Care, Detection of a Meal using CGM. The authors describe a Meal Detection Algorithm that allows them to detect a meal about 30 minutes after the person has started eating!
It's not perfect - but we already know that's true for most of diabetes care. Uf this brings the Artificial Pancreas a little closer and make diabetes treatment a little easier, I'll take it.
Insulin is not a cure. But automatic insulin delivery is a step in the right direction.
Labels: artificial, insulin, JDRF, pancreas
posted by Bernard @ 10:14 AM
4 Comments:
Bernard,
I believe that diabetes will be cured before a safe closed loop system ever comes to fruition.
Riley has had 7 pumps in less than 2 years due to problems with the pump. I just think that a closed loop system will have even more problems.
Would it be great if it happened? I guess. I don't want my son dependant on medical devices. I want a cure.
But, having said that, if there is not a cure but there is a dependable closed loop system then I'll take that...until the cure.
I hope I'm wrong but I just don't see a closed loop system happening. Of course some days I don't ever see a cure happening either.
I would MUCH rather have a cure but until then, sign me up for a closed loop system.
I am willing to try anything at this point.
Hmmm... I'm afraid I remain far from convinced. So this algorithm detects a meal 30 mins after it starts? Insulin begins going in then that won't peak for another 15-30 mins - a full hour after the meal started. Compare that to the 15 min pre-boluses I often give. And what about knowing how big the meal is? An artificial pancreas will only know that either from user input or from the maximum peak of blood glucose level - a maximum peak we all strive to avoid now by judicious timing/sizing of boluses, use of superboluses and combo boluses etc
And... what about algorithms for "Ooo, I just worked out really hard so I'm going to reign in my bolus just a little for this meal" or "I've been really lazy today so I'll actually need a bit extra insulin" etc etc.
The trouble with an artificial pancreas for me is that it will always be reactive, waiting for a blood glucose level rise and then responding. Modern day insulins infused sub-cut (vs intra peritoneal or intravenous) just don't act fast enough to make this viable. Nevermind that an artificial pancreas won't know the factors that led to a high or low, such as stress or a missed bolus that we factor in to our management regularly. One of the big benefits of CGM that so many people seem to miss is that you don't have to just react to a high number with a set dose. You can see if it is already coming down or still going up, and how fast, plus factor in information such as exercise, or a leakage of insulin from an accidentally pulled site, and adjust your standard correction factors accordingly. This may all be possible with an artificial pancreas, but it will require some pretty fancy programs that I think I would have a hard time trusting.
My every day aim right now is to be proactive and head off the rises and falls before they happen, using a combination of my insulin pump, my CGM and the knowledge of my own diabetes that I have inside my head. An artificial pancreas would have to learn fast to pick up the knowledge I've acquired over years and have some pretty smart algorithms to apply the knowledge the way my brain already can.
So no, sorry, I'm not convinced... yet. I'm open to the idea of an artificial pancreas, but need more evidence that it can work as well as I can. Until then, I'd rather go on trusting my head to make these decisions, no matter how much I think doing it sucks.
(No offense is meant to anyone, just my 2pence :-)
I say let's get an implantable chip in our brain and see how that works...or a closed loop system.
Post a Comment
Links to this post:
Create a Link
<< Home