A team of US researchers has developed a novel Artificial Intelligence (AI)-based tool that can pick subtle signs of Alzheimer’s disease that emerge decades before a formal diagnosis is made. The signs are often in the form of irregular behaviors that reflect very early stages of brain dysfunction.
The team from Gladstone Institutes in California engineered mice to mimic key aspects of Alzheimer’s and used the new video-based machine learning tool to detect early signs of the brain disease.
The findings, published in the journal Cell Reports, sheds light on a new strategy for identifying neurological disease earlier than currently possible and tracking how it develops over time.
Gladstone investigator Jorge Palop said that AI can potentially revolutionise how the analysis of Alzheimer’s-linked behaviours — indicative of early abnormalities in brain function — is conducted.
The machine learning platform called VAME, (Variational Animal Motion Embedding) analysed video footage of mice exploring an open arena. It identified subtle behavioral patterns — disorganised behaviour, unusual patterns and transitioning more often between different activities — as the mice aged. These behaviours, likely associated with memory and attention deficits, were captured on camera but may not be noticed by simply looking at the mice.
The tool may help decode the origin and progression of the devastating brain disorders, Palop said, noting it can also be applied to other neurological diseases.
Further, the new study also used VAME to learn whether a potential therapeutic intervention for Alzheimer’s would prevent disorganised behavior in mice.
They found that genetically blocking a blood-clotting protein called fibrin from triggering toxic inflammation in the brain can prevent the development of abnormal behaviors in Alzheimer’s mice.
The team said the intervention also tackled the spontaneous behavioral changes in Alzheimer’s mice.
(Except for the headline, this story has not been edited by The Hindkesharistaff and is published from a syndicated feed.)