A critical time for AI, dementia and better brain health

Bio Beat24 Web Banner

The field of digital solutions for neurodegeneration is exploding. BioBeat24 is an opportunity to get an inside track on the latest developments in diagnosis through to care, fundraising and start-ups. Ahead of the event, Iraida Soria-Espinosa, Senior Innovation & Business Manager at UK DRI, and Miranda Weston-Smith, Founder of BioBeat, caught up with keynote speaker Zoe Kourtzi, Professor of Computational Cognitive Neuroscience at the University of Cambridge.

Zoe Kourtzi

Zoe, we are thrilled you are delivering the keynote, why is finding solutions for dementia so important for you?

The statistics about dementia are overwhelming: over 50 million people worldwide live with dementia, 1 in 2 of us will be affected by dementia in our lifetime, a new patient is diagnosed every 3 seconds. But beyond statistics, dementia touches every family; each one of us has a personal story, a loved family member who battled with dementia and transformed our life perspective for ever.

Yet, we still don’t have sensitive enough tools to diagnose dementia early when new treatments that we now see emerging can really make a difference before damage settles in the brain. We cannot afford to wait, it is now the time to bring together all our creative energy and resources to accelerate discovery and translation towards dementia cure.

What has been your academic journey to Cambridge?

It’s been a long and exciting one, full of surprises! It all started with learning about Pavlov’s classic experiments in a high school in Greece. I thought understanding behaviour was fascinating and quickly switched tracks from Computer Science to Psychology. I finished my undergraduate in Psychology and left beautiful Crete for a PhD in Cognitive Psychology at Rutgers University. I moved to Boston for my first postdoc just as cognitive neuroimaging was emerging- a really exciting time! MIT and then the Max-Plank in Tuebingen, two amazing hubs for cross-disciplinary research gave me the opportunity to flourish as an independent young scientist and taught me how to think out of the box.

Setting up my own lab in Birmingham and now Cambridge has been a wonderful experience, working with the next generation of researchers and inspiring them to reach for grand challenges that tackle real-world problems. In this journey, I’ve been extremely fortunate to be empowered by amazing women - from my beloved aunt who died of dementia to my amazing PhD and postdoc supervisors -and I am truly grateful to them all for their mentorship and guidance every step of the way.

What are your main areas of research?

Cross-disciplinary research is for me the key ingredient for scientific discovery and effective translation and it’s at the heart of our research programme. We bring together expertise from behaviour, neuroscience, computational science and clinical practice to understand the workings of the brain in health and disease. We strive to understand how the brain’s local circuits and global networks interact to support its ability to learn flexibly across the lifespan and plastically reorganise at the face of adversity. We build computational models, harnessing the power of data and AI, to understand the interactive factors that lead to brain and mental health disorders and develop tools that allow us to predict them early and develop interventions tailored to individuals.

I am passionate about developing and translating evidence-based digital tools to clinical practice and confident that working together across disciplines and sectors we have strong potential to change clinical pathways and enhance patient wellbeing.

What are the key challenges and opportunities in applying AI/digital approaches to dementia/neurodegeneration?

Translating AI to dementia has tremendous potential to revolutionise the clinical pathway providing decision support systems for clinicians that allow them to assign the right patient to the right diagnostic and treatment pathway at the right time. Currently, up to a third of patients may be misdiagnosed and others diagnosed too late for treatment to be effective.

Only 2% of patients have access to more predictive tests (PET, lumbar punctures), resulting in misdiagnosis. Improving early prediction and prognosis when treatments can be most effective has strong potential to enhance the wellbeing of patients and their families, while reducing inequalities and costs in healthcare systems. Further, AI has great potential to guide patient stratification for inclusion in clinical trials: tailoring trials to the right patients will accelerate the discovery of new treatments.

Despite the power and promise of AI, the path to translation is not without challenges. We need to build algorithms that are interpretable and interoperable to address the global challenge of dementia. Our tools need to be validated against real-world data from diverse populations and withstand the test of missing and messy data collected routinely in clinical practice.

We need to build AI-guided tools that harness the power of multimodal data to make predictions from non-invasive and less costly data that reduce burden to individuals to deliver early and precise diagnosis and guide personalised interventions. We are now at a critical time: we have all the necessary ingredients (people, algorithms, data); we need to work together and synthesise our resources and creative ideas to build trustworthy digital health solutions for everyone.

Thank you Zoe! We are hugely looking forward to your keynote at BioBeat24.

BioBeat24: Driving solutions in neurodegeneration with digital, takes place on Thursday 16 May, from 13.30-18.15, at the Sainsbury Wellcome Centre, 25 Howland St, London W1T 4JG

For information and to register please visit:

Article published: 25 April 2024