How do we predict?

That's a great question, I'm glad you asked.

PredictBGL builds a customised dynamic mathematical model based on user data to:

  • Calculate personalised insulin doses (using standard insulin dosing algorithms) and;

  • Forecast future blood sugar levels.

The process involves:

  1. Aggregating data from user blood sugar level readings, insulin doses, food and exercise logs, and

  2. Pushing this data through a patented process to determine dosing recommendations and predict blood sugars specific to each user.

The mathematical model tolerates errors relating to accuracy of input data including:
  • Blood glucose meter readings

  • Estimation of carbohydrate intake

  • Dosing ratios

  • Insulin device dosing error

  • Estimation of duration and intensity of factors such as exercise.

If you have further questions, please contact us.

Start now