DIANA develops a personalized, AI-powered digital twin platform to provide adaptive clinical decision support for children and adolescents with Type 1 Diabetes under multiple daily injection therapy.
DIANA stands for Digital twin and Artificial Intelligence for adaptive maNAgement of pediatrics with type 1 diabetes. The project addresses a critical gap: existing decision-support tools for Type 1 Diabetes management are not designed for the unique physiological and lifestyle variability of children and adolescents.
Patient-specific digital twin models built with ReplayBG to simulate individual glucose-insulin dynamics, continuously updated from wearable sensor data.
Multi-component decision support system covering basal insulin dosing, meal bolus management, physical activity recommendations, and safety interventions.
Seamless connection to continuous glucose monitors, activity trackers, and smart insulin pens, feeding real-time data into the platform.
Designed from the ground up for AI-Act, GDPR, and Medical Device Regulation compliance, ensuring safety and explainability throughout.
Rigorous 6-month randomized clinical trial with 60 pediatric participants to validate efficacy and safety of the platform in real-world settings.
Upon project completion, the full platform will be released as an open-source solution to maximize impact and enable adoption by healthcare systems worldwide.
Children and adolescents with Type 1 Diabetes face challenges that adult-focused tools cannot address: rapid growth, puberty-induced insulin resistance, highly variable physical activity, unpredictable meal patterns, and the involvement of caregivers in clinical decisions. DIANA is designed from the ground up around these realities, co-designed with patients, families, and clinicians to ensure it is both clinically effective and practically usable.
DIANA is led by the Bioengineering group at the Department of Information Engineering, University of Padova, in close collaboration with the Pediatric Diabetology Unit of the University Hospital of Padova (AOPU), the single clinical center of the trial.
DIANA follows a structured 48-month roadmap from co-design with stakeholders through clinical validation and open-source release.
Official start of the DIANA project. Research team onboarding, initial co-design preparation, and infrastructure setup.
Completion of participatory co-design sessions with patients, caregivers, and clinical partners to define platform requirements and workflows.
Pediatric-specific digital twin models validated using retrospective datasets. ReplayBG framework adapted for growth and developmental variability.
All decision support system modules (basal, bolus, activity, safety) integrated into the mobile and web platform with full wearable connectivity.
Pre-clinical validation complete. Platform ready for regulatory review and clinical trial preparation.
Ethics committee and regulatory approval obtained for the 6-month randomized controlled trial with 60 pediatric participants.
6-month RCT with 60 children and adolescents with T1D completed. Primary endpoints evaluated and results submitted for publication.
Full DIANA platform released as open-source software. Documentation, datasets, and trained models made publicly available to the research community.
Stay up to date with the latest developments, events, and milestones from the DIANA project.
Loading news…
Scientific publications, conference papers, and software releases produced within the DIANA project.
Loading publications…
For collaborations, open positions, or information about the clinical trial, please reach out to the DIANA team.
DIANA is supported under the FIS 3 (Fondo Italiano per la Scienza) funding scheme, which supports high-impact scientific research projects at Italian universities.