Basket | Login | Register


The VirtualBrainCloud project hosts sessions on AI, dementia and ethics at #30AEC

Thursday 22 October 2020

On 21 and 22 October, the Horizon 2020-funded VirtualBrainCloud (TVB_Cloud) project held two Alzheimer Europe conference sessions on the use of artificial intelligence (AI) in dementia research. The goal of TVB_Cloud is to develop a cloud-based decision support system for clinicians, to help them more accurately diagnose neurodegenerative diseases such as Alzheimer’s and Parkinson’s disease using multi-disciplinary clinical data and personalized brain simulations.   

In the first TVB_Cloud session, chaired by Katarina Stevanovic of TP21, project leader Petra Ritter (Charité Medical University Berlin) presented the TVB_Cloud approach to develop a cloud-based platform for personalized diagnosis and treatment of dementia.  Martin Hofmann-Apitius of Fraunhofer SCAI outlined their knowledge graph approach to increase our mechanistic understanding of neurodegenerative disease, and to identify potentially druggable pathways.  Viktor Jirsa (Aix-Marseille University) rounded off the session, illustrating how algorithms developed to analyse complex systems could help us understand brain networks in health and in disease.

The second TVB_Cloud session addressed the ethical, legal and social issues raised by the use of AI in dementia research.  Bernd Stahl, Ethics Director of the Human Brain Project, outlined how some of the ethical issues raised by the use of AI in healthcare research could be mitigated at policy, organizational and project levels, to ensure that the benefits of AI do not outweigh the risks.  Data protection is a major concern for big data research using AI, and Michael Cepic (University of Vienna) guided the audience through the General Data Protection Regulation, showing how it protects patient privacy whilst enabling health and care research. Rounding off the session, Richard Milne of the Wellcome Sanger Centre addressed the benefits and harms of risk disclosure when risk is detected using algorithm-based approaches, describing a potential route to mitigate harms whilst maximising benefit.