Type 1 - Juvenile Diabetics

Join

T1D Predictive App

Options
okulyd
okulyd Posts: 147 Member
in Social Groups
Hi Everyone,
I'm working on a project for a predictive app. The idea is that the app will suggest things to you such as having a snack or taking a bolus based on foods you eat and activities you do. There are lots of apps out there but this one is different because most don't predict and tell you what to change early enough to prevent going out of range.

My question for this group is, what would you like to see in an app, what should it consider besides food eaten, basal insulin, activity?

How many inputs are too many? I.e. how many questions are you willing to answer before you would throw in the towel and say its too time consuming?

Also if anyone wants to share data with me, I'd love to have some real data such as food, activity, blood sugars and insulin to test the app with.

Also if there is already a great app out there please tell me about it. I am only working on this because I don't think there is one available.

Thanks in advance! Danielle

Replies

  • T1DCarnivoreRunner
    T1DCarnivoreRunner Posts: 11,502 Member
    Options
    The Medtronic CGM has predictive alerts... I would bet Dexcom does also. It tells me when a low or high is predicted based on BG trend.
  • okulyd
    okulyd Posts: 147 Member
    Options
    Trying to build a phone app so not only the T1D is alerted but possibly caregivers are also notified that the T1D might need to take action so they can assist if needed.
  • NovoActive
    NovoActive Posts: 13 Member
    Options
    Hi Danielle

    As mentioned above, to predict when it would likely be necessary to increase/decrease blood sugar levels it would be best to use the live trend in the current blood sugar level as done with CGM's.

    At this point in time, I think it would be difficult to quantify enough of the inputs needed to predict blood sugar level. These would include, but not limited to: BS level, activity (type and intensity), insulin sensitivity and duration of insulin action.

    Personally, I would like a feature that integrates activity + glycemic carbs (total carbs - fibre) + blood sugar readings to keep track how insulin sensitivity changes based on the type of activity for a given state of rest and stress.

    By example:
    A 300 cal walk in not as glycolytic as a 300 cal resistance training squat routine. Thus, more glycemic carbs would be needed to stabilise blood sugar in the case of doing heavy squats relative to walking.

    But I guess this is not one of your main concerns if the app's primary role is to notify caregivers when to take action.

    Regards

Categories