Researchers at Northwell Health’s Feinstein Institutes for Medical Research have developed an AI-based tool that identifies a brain metabolic network capable of predicting psychosis in Alzheimer’s disease (AD).
This discovery offers a potential breakthrough for earlier diagnosis, treatment development and personalised medicine for individuals with AD.
The study, led by Jeremy L. Koppel, associate professor at the Feinstein Institutes’ Institute of Molecular Medicine, and associate professor An Vo, used an AI-derived metabolic brain map that looks at FDG-PET scans of people with AD, including some who experienced psychosis (hallucinations and delusions) and some who did not.
The AI identified a unique pattern of brain activity, like a fingerprint, in the brains of those with psychosis. They called this pattern the ADPN (Alzheimer’s Disease Psychosis Network).
“This discovery of the ADPN provides a critical tool for advancing our understanding and management of psychosis in Alzheimer’s disease,” said Dr. Koppel.
“This biomarker can potentially identify individuals at risk of developing psychosis, allowing for earlier and more targeted interventions.”
Psychosis, characterised by delusions and hallucinations, affects a significant portion of individuals with AD and is associated with accelerated cognitive decline, increased caregiver burden and premature mortality.
Current treatments are limited and often carry significant risks. The ADPN, however, was helpful in distinguishing the difference in brain patterns between people with AD and psychosis, especially between the areas responsible for movement and the areas responsible for language and social understanding.
It was also able to predict who would develop psychosis in the future. The stronger the ADPN “fingerprint,” the worse the person’s cognitive decline.
“The discovery of the ADPN exemplifies the power of AI and interdisciplinary collaboration in advancing our understanding of complex neurological disorders like Alzheimer’s disease,” said Kevin Tracey president and CEO of the Feinstein Institutes and Karches Family Distinguished Chair in Medical Research.
“This innovative approach, integrating neuroimaging with cutting-edge computational tools, offers a new path towards earlier diagnosis, personalised treatment strategies, and ultimately, improved outcomes for individuals affected by this devastating disease.”
Dr. Koppel is an internationally recognised clinical scientist who continues to identify AD’s potential causes and treatments. He recently received a National Institutes of Health grant for nearly $500,000 to look at ptau181 and other biomarkers in psychosis before AD begins as well as a US$1m grant from the Alzheimer’s Foundation of America to expand research into developing new treatments to address the hallucinations, delusions and aggression that come with dementia.
Last year, he and his lab published in the Journal of the American Medical Association (JAMA) Psychiatry papers that showed an increase in specific tau proteins that indicated blood biomarkers could be used to detect psychotic symptoms in AD patient
