Scientists to use AI to analyze 1.6 million brain scans to develop tool that predicts dementia risk

Scientists are to analyse more than a million brain scans using artificial intelligence in a bid to develop a tool that can predict a person’s risk of dementia.

Researchers from the University of Edinburgh and the University of Dundee will examine CT and MRI scans of patients in Scotland, recorded over a decade, as part of a global research project called NEURii.

The team will use AI and machine learning to match the image data with linked medical records, finding patterns that will help doctors better determine the likelihood of someone developing dementia.

The number of people with dementia worldwide is expected to almost triple to 153 million by 2050, and researchers warn this poses a rapidly growing threat to health and social care systems. Health and social costs related to dementia already exceed $1 trillion (£780 billion) a year, research shows.

The scientists want to create digital tools that radiologists can use when scanning patients to determine their risk of dementia and diagnose the disease and related conditions earlier.

They say the tools could also help accelerate the development of more precise treatments for dementia.

Prof Emanuele Trucco, an expert in AI and medical imaging at Dundee, said: “This new dataset will be of great use to neurological researchers.

“Should we achieve a successful proof of concept, we will have a suite of software tools that integrate smoothly and unobtrusively with routine radiology operations, support clinical decision-making and identify dementia risk as early as possible.”

Up to 1.6 million images will be scanned, with approval from the Public Benefit and Privacy Panel for Health and Social Care, part of NHS Scotland.

The data will be held in the Scottish National Safe Haven, which was commissioned by Public Health Scotland to provide a secure platform for the use of electronic NHS data for research.

Project leader, Prof Will Whiteley from the Centre for Clinical Brain Sciences in Edinburgh, said: “Better use of simple brain scans to predict dementia will lead to a better understanding of dementia and potentially earlier diagnosis of its causes, which in turn will make the development of new treatments easier.

“Currently, treatments for dementia are expensive, scarce and of uncertain value. If we can collect data from a large group of people at high risk, who then agree to participate in trials, we can really start to develop new treatments.”