Digital twins are in silico representations of physical systems such as machines or buildings or Formula 1 racing cars. Their use is rapidly becoming everyday practice. However, unlike man-made machines, the complexity and heterogeneity of the human body is such that combining mathematics, computer science, and biology to create a human digital twin is a whole new ball game. We are building a Dietary Digital twin (DDtwin): a versatile, open-source technology platform that combines biology-based (mechanistic, dynamic) and data-driven (machine learning, AI) models, connected to real-life data (sensors, apps) to enable next-gen precision nutrition.
The core of DDtwin is a coupled differential equation-based mechanistic Meal Model, able to accurately simulate whole body glucose and lipid metabolism in response to meals containing carbohydrates, fats and proteins. The Meal Model can serve as a crystallization nucleus for further extension into a comprehensive digital twin.
During the workshop we aim to find potential solutions for a number of technical challenges that need to be overcome in order to extend the Meal Model. These include 1) connecting wearable sensor generated real-time Continuous Glucose Monitoring (CGM) data to the model, 2) linking the dynamic core of DDtwin to models that mimic the gut microbiota (e.g. Genome-Scale Metabolic Models), and 3) design a human-centered user interface for DDtwin that links it to dietary intake data collected with apps. In addition to exploring technical solutions to extend DDtwin, we will make an inventory of ethical, legal, and social issues (ELSI) that need to be addressed to allow meaningful implementation of DDtwin for next-gen precision nutrition.
