VERSION 2.6 IS DEPRECATED IN FAVOR OF CGMSIM v.3⌁

Please head over to CGMSIM v.3 for the new version's documentation !⌁

Welcome to the CGMSIM v2.6 documentation⌁

CGMSIM is an interactive simulation program designed for people without diabetes to experience the effects of different meals, physical activities, and insulin treatments on blood sugar levels in Type I or juvenile diabetes (T1D).

It uses a simulated continuous glucose monitoring (CGM) system to depict changes in blood glucose levels over time, allowing users to better understand how their choices and treatments can affect the condition.

Each simulated adult patient is coupled to a Nightscout website, used to display the CGM curve and to declare meals and insulin doses. Nightscout will be soon provided on request, unless you already have set up your own Nightscout instance.

Here's a quick overview of what it looks like when viewed in fast forward:

New features :⌁

If you opt to use a physical activity tracker, CGMSIM can access your activity data (such as steps and heart rate) from the Fitbit or Google Fit APIs and incorporate them into the simulation to calculate their impact on your simulated patient's blood glucose levels.

CGMSIM can simulate both Multiple Daily Injections (MDI) and Continuous Subcutaneous Insulin Infusion (CSII or insulin pump) treatment regimens. Additionally, users can couple the simulated T1D patient with various DIY artificial pancreas software solutions such as (iOS Loop, AndroidAPS, or openAPS).

Apart from our simplified model, the developers of the LoopInsightT1 simulator simulator have reworked the UVA/Padova simulation mathematics, and we have incorporated this into CGMSIM.

CGMSIM v3.0 is a brand new web application, with a completely rewritten codebase. To use it, you can either create an account or log in using your Google or Fitbit credentials via this link: https://passport-auth-nodejs.herokuapp.com/

\(\textcolor{red}{\textsf{The URL above is temporary !!!}}\)

Email alerts about HIGH and LOW blood glucose have been added. In addition, iOS and Android applications have been created, providing push notifications in the same situations.

CGMSIM now includes optional email alerts for high and low blood glucose levels. Moreover, immediate push notifications for these same situations are possible, using an iOS or Android application.

Introduction⌁

Since even the most accurate physiological simulators can at best provide only an approximation of what happens in a living organism, the goal of this simulation is not even to try to match reality, but to provide a framework for practising, learning and/or teaching.

The treatment goal in type 1 diabetes is to keep the BG or CGM curve values as close as possible to that of a healthy person. In realistic terms, it means 3.9 – 10 mmol/l or 70-180 mg/dl for CGM curves. The Time In Range (TIR) should be at least 70%.

Human physiology⌁

To understand the mechanics of the simulation, a few things must be considered.

1. After a meal, the absorption of carbohydrates from the gut will also increase the blood glucose. Since insulin replacement is administered subcutaneously and not directly into the hepatic portal vein, some of the ingested food will be absorbed into the main bloodstream directly, bypassing the liver. After a short delay, the CGM will reflect the increase of blood glucose.

2. The endogenous glucose production (EGP) is mainly du to the liver release of glucose into the bloodstream. Either by releasing stores (glycogenolysis) or by synthesizing new glucose (neoglucogenesis). Various biological states, alcohol and hormones can affect EGP. As glucose is released into the blood stream, the blood glucose concentration increases.

3. Physical exercise affects blood glucose in many ways. Short bouts of intense physical activity will trigger the release of adrenaline, which increases the EGP. In turn, the increased peripheral uptake of glucose in the muscle tissue will lower the blood glucose as physical activity continues. The same continues also during the post exercise phase, until the muscle glycogen stores are repleted.

4. Mealtime insulins have a short duration of activity (DIA, ~3 hours), while long-acting insulin analogs are meant to be injected once or twice a day. Both will lower blood glucose and thus CGM values.

5. The intricate biological pathways affecting the blood glucose variations are difficult to replicate mathematically. The absorption of ingested food from the gut, the absorption of insulin from the subcutaneous tissue, the timing of exercise, the circadian rhythm, menstrual cycle, the moon phase and the alignment of planets and stars 😉 all seem to play a role. Adding a touch of random variation (or chaos) brings some realism in the simulation.

Why use a simulator ?⌁

Many scenario-based simulators have been written in the past, like AIDA or the UVA/Padova T1DMS. In contrast, CGMSIM runs in real time like a "serious game". It requires time and dedication to reach its goal. It is not a matter of hours or days, but rather weeks or months.

This website will describe step by step how create your virtual subject's profile, and how to use the simulator.

Where is the source code ?⌁

The source code and more technical details on how various scripts are excecuted and interact with each other can be found in the CGMSIM project's repository