Researchers from the University of Cambridge’s department of psychology have revealed that their artificial intelligence tool for predicting early Alzheimer’s disease (AD) more accurately detected the condition compared to current clinical tests.
Published in eClinical Medicine, the machine learning model was able to predict whether and how fast an individual with mild memory and thinking problems will go on to develop the neurological disease.
Affecting more than 944,000 people in the UK, dementia is a neurological condition that affects the ability to remember, think or make decisions in everyday life.
Recognised as the most common form of dementia, AD accounts for between 60% and 80% of all cases.
Around a third of patients are often misdiagnosed or diagnosed too late for treatment to be effective due to invasive or expensive tests, including positron emission tomography (PET) scans or lumbar punctures, which are not available in all memory clinics.
Researchers used routinely collected, non-invasive and low-cost patient data from over 400 individuals as part of a research cohort in the US to build the model and then tested it on real-patient data from an additional 600 participants from the US cohort, along with longitudinal data from 900 people from memory clinics in the UK and Singapore.
The algorithm successfully distinguished between people with stable mild cognitive impairment and those who progressed to AD within a three-year period, while also correctly identifying individuals who developed AD in 82% of cases and those who did not in 81% of cases from cognitive tests and an MRI scan.
In addition, the algorithm was around three times more accurate at predicting progression to AD than the current standard of care, including standard clinical markers or clinical diagnosis, which was checked using follow-up data over six years, proving its potential in reducing the chances of misdiagnosis for patients, which researchers say could be applicable in a real-world patient clinical setting.
Senior author professor Zoe Kourtzi, department of psychology, University of Cambridge, commented: “This [tool] has the potential to significantly improve patient well-being” and “will also help remove the need for unnecessary invasive and costly diagnostic tests.”