AI and Alzheimer's: how the Turing is pioneering new research into this devastating disease

Tuesday 21 Sep 2021

Today Tuesday 21 September, is World Alzheimer’s day. Part of World Alzheimer’s Awareness month, it is an international campaign created to raise awareness, educate and highlight issues faced by people affected by dementia. This year the Alzheimer’s Society is spreading the message: Knowledge is Power. With nearly 50 million people living with dementia worldwide, this global issue can only be tackled through collaboration and working together.

The Alan Turing Institute is committed to using our resources to find innovative ways, using artificial intelligence (AI) and data science, to diagnose, prevent and treat Alzheimer’s. This is demonstrated in our partial funding of the Deep Dementia Phenotyping (DEMON) Network; an international network of academics, clinicians and other partners, seeking to revolutionise dementia research and healthcare by harnessing the power of AI and data science. The Turing is also partnering with Alzheimer’s Research UK and others, to lead the Analytics Hub for EDoN – Early Detection of Neurodegenerative Diseases. Our data scientists are providing the platform that will analyse the data collected and will develop, validate and refine machine learning ‘fingerprint’ models that can detect the diseases apparent in the earliest stages of dementia. To mark World’s Alzheimer’s day, EDoN has published a report looking at public attitudes to dementia, including attitudes to technology and wearables for risk and diagnosis.

A recent success includes trialing an AI system to diagnose dementia after a single brain scan. Researchers, led by Professor Zoe Kourtzi, a Fellow at the Institute, have developed an algorithm which can spot structural changes within the brain that happen in the early stages of Alzheimer’s. Whilst still in a primary phase, this trial shows encouraging results in securing an early diagnosis of the disease. An early diagnosis can allow researchers to develop new medications and treatments, and support clinicians to intervene early with potentially better outcomes.

As the UK’s national institute for data science and AI we have created an opportunity for talented multi-disciplinary researchers to join intensive ‘collaborative hackathons’ through our popular Data Study Groups (DSG). Our researchers act as ‘challenge owners’ and are provided with real-world datasets to brainstorm and engineer data science solutions. They then have the opportunity to present these back to industry, government and third sector organisations. This month’s DSG has a focus on dementia, with challenges ranging from modelling amyloid beta plaque formation in Alzheimer's disease to using machine learning to improve sleep habits in Dementia patients.

Whilst the Turing have made some significant leaps in finding way to better diagnose and treat Alzheimer’s, there is still more work to do. With an ageing population, by 2040 it is projected that 1.6 million people will be living with dementia in the UK. World Alzheimer’s month provides a stark reminder that more research, funding and collaborative work must be done to support those living with, and caring for, people with Alzheimer’s.