It’s been used to detect eye diseases, make medical diagnoses, and spot early signs of oesophageal cancer. Now it has been claimed artificial intelligence may be able to diagnose dementia from just one brain scan, with researchers starting a trial to test the approach.

The team behind the AI tool say the hope is that it will lead to earlier diagnoses, which could improve outcomes for patients, while it may also help to shed light on their prognoses.

Dr Timothy Rittman, a senior clinical research associate and consultant neurologist at the University of Cambridge, who is leading the study, told the BBC the AI system is a “fantastic development”.

“These set of diseases are really devastating for people,” he said. “So when I am delivering this information to a patient, anything I can do to be more confident about the diagnosis, to give them more information about the likely progression of the disease to help them plan their lives is a great thing to be able to do.”

It is expected that in the first year of the trial the AI system, which uses algorithms to detect patterns in brain scans, will be tested in a “real-world” clinical setting on about 500 patients at Addenbrooke’s hospital in Cambridge and other memory clinics across the country.

“If we intervene early, the treatments can kick in early and slow down the progression of the disease and at the same time avoid more damage,” Prof Zoe Kourtzi, of Cambridge University and a fellow of national centre for AI and data science the Alan Turing Institute, told the BBC. “And it’s likely that symptoms occur much later in life or may never occur.”

Dr Laura Phipps at Alzheimer’s Research UK said Kourtzi was also leading a research project, funded by the charity, that used data from wearable technology to predict diseases like Alzheimer’s 15-20 years earlier than it was currently possible. Phipps added that the application of AI to brain scans might bring benefits.

“To diagnose dementia today, doctors need to rely on the interpretation of brain scans and cognitive tests, often over a period of time,” she said. “Machine learning models such as those being developed by Prof Kourtzi could give doctors greater confidence in interpreting scans, leading to a more accurate diagnosis for patients.”

Phipps added that it is hoped such approaches may eventually help to detect the diseases that cause dementia much earlier.
“This would have a huge impact on people with dementia and their families,” she said.

However Prof Tara Spires-Jones, deputy director of the Centre for Discovery Brain Sciences at the University of Edinburgh, who was not involved in the study, said excitement might be premature.

“Finding ways to diagnose dementias very early in the disease process is a very important goal that will help both research and eventually treatment, but it looks like this is still in fairly early stages,” she said.

Prof Clive Ballard, a dementia expert at the University of Exeter, agreed. “AI has been shown to improve the diagnostic potential of brain scans compared to clinical reading of the scans, but there is so much heterogeneity between individuals that it is completely infeasible for a single scan, biomarker or clinical test to be that certain in a single assessment,” he said.

“This approach is definitely a positive direction of travel that will lead to improvements in diagnosis, but we need to be really careful not to create false expectations.”