Products You May Like
Scientists in the United Kingdom are working to train an artificial intelligence model to recognize signs of dementia in brain scans. They hope their tool will someday offer insight into individuals’ risk of experiencing dementia, as well as trace the development of a person’s cognitive decline after it has already begun.
The team, which includes 20 researchers from the University of Edinburgh and University of Dundee, received approval from the UK’s National Health Service (NHS) to leverage a trove of brain scans for the AI’s development. The dataset includes 1.6 million computed tomography (CT) and magnetic resonance imaging (MRI) scans, which detail the brain’s structure at a given point in time. Each scan belongs to a patient in Scotland who underwent imaging for a cognitive concern between 2008 and 2018.
Because CT and MRI snapshots reveal physical degradation thought to be linked to cognitive decline, training an AI model on these scans could allow radiologists to spot subtle signs of dementia in future patients. With roughly one in three cases of dementia considered preventable, identifying the warning signs of cognitive decline could allow otherwise healthy patients to improve their vascular health by quitting or avoiding smoking, becoming more physically active, or keeping their blood pressure within a healthy range—all factors researchers believe contribute to increased risk of dementia. For certain forms of dementia, like Alzheimer’s disease, experimental medications could also slow disease progression.
The AI tool, called Scottish AI in Neuroimaging to predict Dementia and Neurodegenerative Disease (SCAN-DAN), will be overseen in part by Edinburgh Innovations, the University of Edinburgh’s research and technology commercialization service. SCAN-DAN is one of three programs comprising NEURii, a brain health-focused research collaboration between several educational institutions, think tanks (including Bill Gates’ Gates Ventures), and healthcare entities. Will Whiteley, a neurologist at the University of Edinburgh, says this collaboration is what will empower SCAN-DAN to become a real tool instead of being used purely for experimentation or research purposes.
SCAN-DAN joins a number of efforts to “catch” dementia risk factors or early-onset Alzheimer’s disease. This year, two different organizations introduced their own Alzheimer’s blood tests, which are said to identify signs of Alzheimer’s more comfortably and with a higher level of accuracy than spinal taps, cognitive screenings, and even the types of brain scans involved in SCAN-DAN. Researchers in the United States also introduced last year an electroencephalography (EEG) that uses brain waves to detect early signs of Alzheimer’s.