About AI-Mind research
The goal of the research is to reduce the disease’s burden by developing novel, artificial intelligence(AI)-based tools to support healthcare professionals in their diagnosis and offering timely interventions to patients.
Besides time-consuming patient investigations with low discriminative power for dementia risk, current treatment options focus on late symptom management. What is now complex, labour-intensive, costly, and poorly predictive screening methods for mild cognitive impairment (MCI) shall be replaced by automated diagnostic screening tools. These are driven by artificial intelligence to address the urgent need for early accurate diagnosis and risk prediction.
AI-Mind will disrupt the clinical setting and significantly impact patients’ and doctors’ diagnostic journeys. Physicians will have a supportive decision-making tool that can identify patients at risk of developing dementia with a high probability.
With the currently available technology, many patients receive their diagnosis only after the onset of dementia. As a consequence, there might not be the opportunity to start preventive therapies.
For people with MCI, the dementia risk is almost 30% higher than unaffected individuals. Therefore, we need effective diagnostic tools for early dementia risk assessment and intervention for people with MCI.
Activities in the AI-Mind are organised in nine integrated work packages (WPs). These WPs are focusing on different stages and aspects of developing the AI-based tools, specifically:
- concept governance and assessment of clinical landscape for tools implementation
- data management and feature extraction
- AI-modelling for the AI-Mind Connector and AI-Mind Predictor
- Data analysis and visualisation platform for AI-Mind tools
- clinical implementation and validation of AI-Mind Connector and AI-Mind Predictor
- health technology assessment of AI-Mind tools
- project outreach and exploitation of results
- innovation management
- ethic requirements
An innovative solution for early detection of dementia
At the heart of AI-Mind are two AI-based digital tools that will be integrated into a cloud-based diagnostic support platform. These two tools will process routinely collected data in an innovative way.
The AI-Mind Connector, which is fed with brain images from electroencephalography (EEG) data, will evaluate and visualise interactions between different brain areas, identifying early disturbances in the functional brain network. AI-Mind reads these disturbances as the first signs of cognitive changes that may develop into dementia in the future.
The AI-Mind Predictor, which makes use of AI to combine data from the Connector, cognitive tests and blood analysis, will provide an accurate (>95%) prediction of dementia risk for clinical decision making.
Both tools will be integrated into a cloud-based diagnostic platform, providing an easy-to-implement service for health professionals.
The dementia patient journey at the moment involves several clinical visits, which finally result in a diagnosis of dementia two to five years after initial symptoms.
AI-Mind will shorten the time to diagnosis to a week and allow for diagnosis at the MCI stage, when there are no severe structural cerebral defects and intervention is still possible. AI-Mind will extend the ‘dementia-free’ period among MCI-patients who are prodromal to dementia by offering diagnosis and early intervention.
AI-Mind is an European project receiving funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 964220. It is a five-year Research and Innovation Action (RIA) that officially starts in March 2021. It involves fifteen project partners from eight European countries and has a budget around 14 million euro.
The AI-Mind project addresses the Horizon 2020 Call H2020-SC1-BHC-06-2020, ‘Digital diagnostics – developing tools for supporting clinical decisions by integrating various diagnostic data,’ by developing a decision support diagnostics platform for early risk estimation of dementia through two BSOA AI-based tools using SOA data.