• Guest - w'd love to know what you think about the forum! Take the 2025 Survey »

More cheerful stuff (really this time)

I think the term 'artificial pancreas' is a misnomer. I agree that the ability for pumps to read bsls and be programmed to inject insulin in response is wonderful, but the gadget cannot detect how many carbs and protein has been consumed and digested by the user.. That has still to be entered in by the user.
Unless the pump uses insulin injected under the skin which is responsive enough to control BSL changes quickly enough the control will be likely always be inferior to what real, fully functional pancreases can do.
Just my 10 p worth (+ VAT).
 
Our very own @tim2000s has been working very hard on closed loop systems for a while now.
 
I've been using a closed loop for over two years now. With Fiasp, when it works, it's possible to eat meals of up to 50g without needing to announce the meal or enter a bolus. I'm quietly confident that the new formulations from Lilly will make that even better.

It's great, and means I maintain around 85%-90% time in range (clinical trial range 70-180 mg/dl, 3.9-10 mmol/l) and have to do very little to manage type 1.

But then, we've had an Open Source group working on the algorithm and making it do what we want for around three to four years now.
 
I've been using a closed loop for over two years now. With Fiasp, when it works, it's possible to eat meals of up to 50g without needing to announce the meal or enter a bolus. I'm quietly confident that the new formulations from Lilly will make that even better.

It's great, and means I maintain around 85%-90% time in range (clinical trial range 70-180 mg/dl, 3.9-10 mmol/l) and have to do very little to manage type 1.

But then, we've had an Open Source group working on the algorithm and making it do what we want for around three to four years now.



Just way too clever for mere mortals like me.
 
I've been using a closed loop for over two years now. With Fiasp, when it works, it's possible to eat meals of up to 50g without needing to announce the meal or enter a bolus. I'm quietly confident that the new formulations from Lilly will make that even better.

It's great, and means I maintain around 85%-90% time in range (clinical trial range 70-180 mg/dl, 3.9-10 mmol/l) and have to do very little to manage type 1.

But then, we've had an Open Source group working on the algorithm and making it do what we want for around three to four years now.
That’s incredible stuff, I’d love to do looping but until someone works out a hack for Omnipod....

Are you using a Raspberry Pi sort of setup?
 
The one which made me both laugh and admire the chutzpah of it was the bloke who fitted his openAPS rig inside a tic-tac box!

C38lN11WMAERZaV.jpg
 
The one which made me both laugh and admire the chutzpah of it was the bloke who fitted his openAPS rig inside a tic-tac box!

View attachment 28809
That's pretty normal:
IMG_1336.jpg


@tim2000s how do the rest of us get one of these?
If you're interested. start here: https://www.diabettech.com/looping-a-guide/

You can now loop with one of:
  • Old medtronic pump
  • Roche Combo
  • Roche Insight
  • Dana RS
That’s incredible stuff, I’d love to do looping but until someone works out a hack for Omnipod....

Are you using a Raspberry Pi sort of setup?

Similar - I use an Edison and Explorer board. If you build AndroidAPS, you can use it open loop, so it will make suggestions to help you keep aligned, and allows you to use the IOB aware bolus wizard.
 
All of this is beyond me but I'd love to understand and have one!!
 
I have enough bother pairing my Fitbit and phone!

Oh well...........
 
This was on Trust Me.. a few months ago but the team at Cambridge were on Radio 5 this morning sounding pretty confident. It seems to work and patients report ;not feeling like a diabetic' with parents of type 1s finally getting a good night's sleep. So definitely good news though right now I'd just like the FSL that I am owed!
 
I think the term 'artificial pancreas' is a misnomer. I agree that the ability for pumps to read bsls and be programmed to inject insulin in response is wonderful, but the gadget cannot detect how many carbs and protein has been consumed and digested by the user.. That has still to be entered in by the user.
Unless the pump uses insulin injected under the skin which is responsive enough to control BSL changes quickly enough the control will be likely always be inferior to what real, fully functional pancreases can do.
Just my 10 p worth (+ VAT).
Oh I thought it very much did detect the blood sugar every 10 mins then used an algorithm to adjusts the dose. Do not know if it learns on the job though. Agree that manual entry is not foolproof as in I can do that now but am bad at it and there are too many variables...
 
Back
Top