LLMs are AI models capable of reading, understanding and analysing human language. This enables them to process large volumes of care information more quickly and identify connections that are less obvious to humans. The project focuses on two specific challenges: recognising changing care needs at an earlier stage and reducing the administrative burden on healthcare professionals.

SCOPE is a collaboration between healthcare organisations, knowledge partners and innovation partners. Organisations including SDW, SOVAK, MindLabs, the Care Innovation Centre, Avans University of Applied Sciences and Tilburg University are working together to develop and test this technology.

Project 1: Identifying changing care needs

Key question: How can LLMs help to identify changes in clients’ behaviour, wellbeing and health at an earlier stage?

What are care needs?

In long-term care, clients’ care needs are constantly changing. However, these changes are not always noticed straight away. Daily observations often contain valuable signals, but due to work pressure and the volume of reports, patterns sometimes remain hidden.

In this project, we are investigating how LLMs can analyse this unstructured data and make subtle changes visible more quickly. This will enable care teams to intervene earlier and better tailor care to the client’s current situation.

Intended outcomes

• Faster identification of changing care needs
• Better and more proactive care interventions
• Greater insight into client development
• Higher quality of care

Project 2: LLMs for administrative support

Main question: How can LLMs support healthcare professionals in processing and structuring client information?

Support

Healthcare professionals spend a great deal of time manually processing information from medical records and initial consultations. This takes up time that cannot be spent on direct care.

As part of this project, we are investigating how Large Language Models (LLMs) can automatically identify and summarise relevant information and incorporate it into the correct sections of client records. This will streamline administrative processes and free up more time for personalised care.

Expected outcomes

• Reduced administrative burden
• Faster and more consistently completed client records
• Lower risk of errors in data processing
• More time for direct care

Approach and methodology

In both projects, healthcare data is collected, cleaned and analysed. Researchers then train and test various language models in collaboration with healthcare professionals.

Thanks to this co-creation, the applications remain not merely theoretical but directly linked to practice. In this way, we are building AI solutions that truly reflect the day-to-day reality of healthcare.

From September 2026, a Futures Lab team will begin working on this project to turn these initiatives into reality.

Contact

For further information, please contact
Robin den Oudsten
Robin.denoudsten@mind-labs.eu