An algorithm to determine Alzheimer’s disease

Machines can predict Alzheimer’s disease up to two years in advance.

The futuristic idea isn’t as far-fetched as it sounds. Researchers from McGill University in Montreal, Quebec, Canada, built an algorithm to predict who would develop dementia. Their findings have been published in the journal Neurobiology of Aging.

Using artificial intelligence techniques and big data, researchers from McGill’s Douglas Mental Health University Institute Translational Neuroimaging Laboratory developed an algorithm capable of recognizing dementia from a single amyloid positron emission tomography (PET) scan of the brain. The algorithm had 84 percent accuracy.

“By using this tool, clinical trials could focus only on individuals with a higher likelihood of progressing to dementia within the time frame of the study,” says Serge Gauthier, CM, MD, FRCPC, co-lead study author and professor of neurology and neurosurgery and psychiatry at McGill in a press release. “This will greatly reduce the cost and the time necessary to conduct these studies.”

Researchers focused on people with mild cognitive impairment (MCI), a condition that causes accumulation of amyloid and can lead to dementia. They taught the algorithm to identify who would develop dementia using hundreds of PET scans of people with MCI.

Medications to delay disease onset need to be tested in clinical trials that run between 18 and 24 months to determine medication effectiveness. However, it can be challenging at best to predict who will develop the disease.

AI could be a helpful diagnostic tool to identify people at greater risk for developing dementia. Scientists would have access to better candidates for clinical drug trials that could potentially lead to more accurate findings and new treatments.

Researchers are now working to see if other dementia biomarkers could be incorporated into the algorithm to improve prediction capabilities. They are also testing the algorithm in different patient cohorts. The have made the software available for other scientists online at https://predictalz.tnlmcgill.ca/PredictAlz_Amy.  

Nicole Stempak

Nicole was Senior Editor at I Advance Senior Care and Long Term Living Magazine 2015-2017. She has a Journalism degree from Kent State University and is finalizing a master’s degree in Information Architecture and Management. She has extensive studies in the digital user experience and in branding online media. She has worked as an editor and writer for various B2B publications, including Business Finance.