Data streams
Simulated continuous glucose monitoring and personal health-record data provide the foundation.
The project · 116911
Decision support for a more personalized approach to diabetes.
MediBRAIN combines artificial intelligence, glucose data, wearables, and personal health records to explore how technology can support better-informed health decisions.
What we are building
A bridge between complex data and everyday understanding.
The project develops an innovative system for predicting blood-glucose levels and presenting meaningful information through the existing MedInfoBook health platform.
It brings together continuous glucose monitoring, machine-learning models, digital twins, and interoperable health-data standards in one modular environment.
How it works
Four building blocks.
One connected system.
Machine learning
Personalized models forecast glucose levels up to 30 minutes ahead.
Digital twins
Twelve anonymized digital twins support development, testing, and evaluation.
Interoperability
FHIR R4 resources and modular APIs connect the system to the wider health ecosystem.
The current implementation is a proof of concept and pilot system. It uses simulated data and digital twins, is not a clinical trial, and is not certified as a medical device or intended for clinical decision-making.
At a glance
12digital twins
30 minprediction horizon
100%valid FHIR R4 resources
8.5/10integration evaluation