Explaining Artificial Pancreases - a catch all topic

[slip]

@tim2000s I do love your graphs, couple of questions, do you wear an activity tracker? (and my safe bet is yes!) - I'd love to see that overlaid on these.

Could you talk us through and explain those to days in a bit more detail - Monday at 08:10 you had a bolus of .7u, 8:45 4.6u then had breakfast at 09:45 and had 18g carbs and 1.8u - what was that all about? pre-bolasing I assume, same with lunch?

Also what happened in the wee hours of Tuesday? What's the significance of the light blue and darker mid-blue basals?

 

[tim2000s]

Following a recent comment, this is a topic to discuss behaviours seen in hybrid closed loop systems, give examples, etc, for those who are interested. We'll start with this view of a couple of days in October, then look at the questions that have been asked.

 

[tim2000s]

Could you talk us through and explain those to days in a bit more detail - Monday at 08:10 you had a bolus of .7u, 8:45 4.6u then had breakfast at 09:45 and had 18g carbs and 1.8u - what was that all about? pre-bolasing I assume, same with lunch?

Probably not that easy as these graphs are from nearly two months ago, but I'll dig through my history and walk you through what's going on.

Also what happened in the wee hours of Tuesday? What's the significance of the light blue and darker mid-blue basals?

That's a compression low, where you lie on the sensor and it shows low, without there really being one. They show up as a rapid drops and fairly rapid recoveries.

 

[Engineer88]

can I ask what software you use? it looks like diasend only I cant get my basal like you do (or I'm using something wrong?)

 

[slip]

That's a compression low, where you lie on the sensor and it shows low, without there really being one. They show up as a rapid drops and fairly rapid recoveries.

I thought that was a possibility but as the AP is thinking you're low, suspends or reduces the basal I would have thought you'd have gone a lot higher once the compression had past, but that doesn't seem to be the case.

 

[tim2000s]

I thought that was a possibility but as the AP is thinking you're low, suspends or reduces the basal I would have thought you'd have gone a lot higher once the compression had past, but that doesn't seem to be the case.

It only really amounted to 30mins without insulin, and once it detects a rise, it turns insulin back on, so even if you did go up a bit, it wouldn't be massive.

 

[tim2000s]

can I ask what software you use? it looks like diasend only I cant get my basal like you do (or I'm using something wrong?)

This is NightScout. http://www.nightscout.info/

 

[tim2000s]

@slip Hope this makes some things a bit clearer.

 

[slip]

as mud - lol!

actually yes it helps, you can clearly see it reacting to a steep drop and suspending basal, and you take about an hour to react to it and start to climb back up - @ 18:10 following you unneeded correction, and then again at 20:30 - that's the prediction algorithm working!

Lost some of the text regarding bolousing

 

[slip]

@donnellysdogs

 

[tim2000s]

@slip the blue bolus model is described here: https://diyps.org/tag/eating-soon-mode/

Essentially, around an hour before eating, you set a temporary target of 4mmol/l and give a small pre-bolus that would bring you down to 4.0 from where you currently are. The idea is that this additional bolus "activates" glut4 transporters and gets you ready to absorb carbs.

Before your meal you do the meal bolus less the earlier pre-bolus and go from there. If you use the Advanced Meal Assist function then you give around 60% of total required meal bolus, and the AMA system calculates your carb absorption and adjusts your insulin from there. I've found it works very well.

 

[slip]

I heard about pre-blousing like that before (it would only fit my lunch time schedule anyway) - but I presume the AP would pick up that the meal wasn't forth coming and deal with it by dropping basal. The AMA sounds really good.

 

[tim2000s]

I heard about pre-blousing like that before (it would only fit my lunch time schedule anyway) - but I presume the AP would pick up that the meal wasn't forth coming and deal with it by dropping basal. The AMA sounds really good.

That's what happens, yes. Basically it looks at your IOB vs COB and adjusts.