Personalising Diabetes Care in Europe and Beyond: The MELISSA Project Launches Clinical Trial

24 November 2025 – For the first time, MELISSA brings together two AI engines in one platform: automated, real-time dietary assessment and self-learning advisor for insulin dosing, enabling fully personalised therapy. The EU-funded research project MELISSA aims to innovate diabetes management in Europe and beyond by making expert-level insulin adjustment recommendations accessible to all. Following the completion of the in-silico simulations, successful integration, and feasibility-study validation of the insulin recommendation algorithm earlier this year, the platform is now ready for practical testing in a clinical trial.

Transforming Personalised Insulin Therapy

The MELISSA app is designed to offer people with diabetes personalised insulin intake guidance, enabling them to make more knowledgeable treatment decisions, significantly enhancing their overall quality of life. This innovative way of personalising insulin use relies on a comprehensive self-learning system – ABBA algorithm – that continuously learns and adapts to optimise daily insulin dosing based on individual needs, independent of the diabetes management technology in use. The mobile app integrated into the platform combines real-time dietary assessment and optimised insulin dosing, while personalising recommendations based on individual eating habits. Additionally, the platform can work with any type of glucose monitoring device, making it accessible to a wide range of users and healthcare providers. Extensive in-silico validation – which means using computer simulations for tests – has ensured the platform’s accuracy and effectiveness in personalising insulin therapy for each user.

Clinical Trial Across Four European Countries

With the platform now completely developed and validated the next step is the clinical trial – a crucial milestone toward bringing this transformative technology to a real-world clinical setting. The clinical trial will be conducted as a multicentre study across four countries: Denmark, Germany, Greece, and the Netherlands. The trial will include approximately 500 people with type 1 and type 2 diabetes on multiple daily injections with insulin. Over a period of 22 weeks the MELISSA platform will be used by both individuals with diabetes and healthcare providers, helping to refine and expand its practical applications.

AI-Driven Diabetes Management

Currently there is no smartphone app combining both AI-based insulin recommendations and an automatic dietary assessment system, offering CHO estimation, available for the wider diabetes community. “MELISSA will offer individuals with diabetes and healthcare providers a tool for effective, data-driven diabetes management,” elaborates Bastiaan de Galan, coordinator of the MELISSA project. In addition to validating the algorithm and launching the clinical trial, the Swiss partners of the MELISSA consortium conducted a feasibility study, which further showed that real-world use of ABBA follows the same positive trajectory observed in the in-silico simulations; furthermore, the consortium is investigating diabetes self-management skills with the EDUCATE study. This brings the project one step closer to improving diabetes care across Europe and beyond for a better quality of life for people living with diabetes through validated innovative AI applications.

About MELISSA

The MELISSA project strives to fill the gap between AI and its validated application in daily diabetes management routine by introducing the MELISSA platform. The platform will comprise two applications: a mobile app for people living with diabetes and a web service for healthcare providers. By moving toward a future where technology, AI, and personalised medicine converge, MELISSA aims to make diabetes management more effective, individualised, and accessible.

Contact
Project Coordinator:
Bastiaan de Galan
Universiteit Maastricht, The Netherlands