Digital tools for screening brain connections and estimating the risk of dementia in people with mild cognitive problems. That is what is being worked on within the international AI-Mind project. Radboudumc, among others, is participating in the project, which is led by Oslo University Hospital.
Every three seconds, someone somewhere in the world develops dementia. Over 50 million people worldwide are affected by the disease. This number may even rise to more than 80 million by 2030. The impact of the disease is huge. For instance, patients' independence comes under pressure. In many cases, people with dementia are diagnosed late, which reduces their chances of deciding their own future and receiving the right care to remain independent for as long as possible.
Time-consuming and not always reliable
Diagnosis is currently not only time-consuming, but also not always reliable. In addition, treatment is mainly focused on late symptoms of the disease. Dementia leads to significant costs for patients' families, medical care and society. By pushing dementia back to a later stage in life, the studies hope to significantly reduce these costs.
Two new tools based on artificial intelligence (AI) are being developed to achieve this goal: the AI-Mind Connector and the AI-Mind Predictor. The AI-Mind Connector is aimed at diagnosing dementia and identifies dysfunctional brain networks. To do so, the tool relies on EEG scans, among other things.
Based on data collected with the AI-Mind Connector, the AI-Mind Predictor estimates the likelihood of developing dementia. In doing so, the tool also looks at genetic tests and digitised cognitive tests.
Intelligent diagnostic platform
Together, the two tools form an intelligent diagnostic platform. The platform provides a report for individual patients, which includes recommendations for treatment. The pair of instruments should thus replace the labour-intensive, costly and poorly predictive screening procedures currently used. By using AI, the tools can automate these procedures.
"In fact, AI-Mind is developing a tool that predicts which people with mild cognitive problems will develop dementia," says Tim Govers, who is involved in the project from Radboudumc. His contribution lies mainly in Health Technology Assessment (HTA), in assessing this new health technology.
Govers: "We are evaluating the effects of this tool for patients and healthcare. We already do this during the development of the instrument, so that adjustments can be made early on where necessary. Furthermore, we use decision models to see how this instrument can be used as cost-effectively as possible in different countries and health systems."
Diagnosing faster
Among other things, the researchers hope to use their platform to significantly reduce the time it takes to make a diagnosis. Whereas such a diagnosis now sometimes takes several years, the platform should reduce this to a week. The faster analysis gives the patient and doctor more time for preventive interventions, with the aim of further reducing disease onset. For example, the researchers point out that people with mild cognitive impairment are 30 per cent more likely to develop dementia than average.
In addition to Radboudumc and Oslo University Hospital, various other hospitals, universities and organisations are also taking part in the project. An overview of all the partners involved is available here.
The AI-Mind project is a Horizon 2020 project. This is a financial instrument through which the European Commission supports innovative projects. The instrument is designed to secure the European Union's global competitiveness.
Author: Wouter Hoeffnagel
