Scientists plan to use artificial intelligence to analyse more than one million brain scans in a bid to develop tools to predict dementia risk.
Researchers from the Universities of Edinburgh and Dundee will examine CT and MRI scans of Scottish patients collected over a decade as part of a global research effort called NEURii.
The research team plans to use AI and machine learning to match the imaging data with relevant health records to find patterns that can help doctors more accurately assess a person's risk of developing dementia.
The number of people living with dementia is predicted to nearly triple to 153 million worldwide by 2050, and researchers warn that this poses a fast-growing threat to health and social care systems. Research shows that health and social care costs associated with dementia already exceed $1 trillion (£780 billion) a year.
Scientists aim to develop digital tools that radiologists could use to scan patients to determine their risk of dementia and diagnose the disease and related symptoms earlier.
The tool could also help speed up the development of more precise treatments for dementia, the researchers said.
Professor Emanuele Trucco, an expert in AI and medical imaging at the University of Dundee, said: “This new dataset will be extremely useful for neurology researchers.”
“If the proof of concept is successful, we will have a suite of software tools that will integrate smoothly and unobtrusively into everyday radiology practice to support clinical decision-making and warn of dementia risk as early as possible.”
Up to 1.6 million images will be scanned following approval from the Public Service and Privacy Commissioner for Health and Social Care, part of the Scottish NHS.
The data will be stored at Scotland's National Safe Haven, which is commissioned by Public Health Scotland and provides a secure platform to make NHS electronic data available for research.
Project co-leader Professor Will Whiteley, from the Edinburgh Centre for Clinical Brain Sciences, said: “Better use of simple brain scans to predict dementia could improve our understanding of the condition, enable earlier diagnosis of its causes and in turn make it easier to develop new treatments.”
“Currently, dementia treatments are expensive, few in number and of uncertain value. If we can collect data from large numbers of people at risk and get them to agree to take part in clinical trials, we can really start to develop new treatments.”
