New AI tool help detect Alzheimer’s linked behaviours decades early

Researchers in the US have developed a groundbreaking Artificial Intelligence (AI) tool capable of identifying subtle signs of Alzheimer’s disease decades before clinical diagnosis. These early indicators often appear as irregular behaviors linked to the initial stages of brain dysfunction.

A team from the Gladstone Institutes in California engineered mice to mimic key aspects of Alzheimer’s disease and utilized the AI-powered, video-based tool to detect early-stage symptoms. The findings, published in Cell Reports, introduce a novel approach for diagnosing neurological diseases earlier and monitoring their progression over time.

The machine learning tool, named VAME (Variational Animal Motion Embedding), analyzes video footage of mice in an open arena to detect subtle behavioral changes. These include disorganized actions, unusual movement patterns, and frequent transitions between activities, which are often linked to memory and attention deficits. While these behaviors may not be evident to the human eye, VAME can identify them with high precision.

Jorge Palop, a Gladstone investigator, noted that AI has the potential to revolutionize the analysis of Alzheimer’s-related behaviors, providing valuable insights into early brain abnormalities. Additionally, the tool could be applied to study other neurological disorders, expanding its impact beyond Alzheimer’s research.

The study also explored therapeutic interventions for Alzheimer’s using VAME. Researchers found that blocking the blood-clotting protein fibrin, which triggers toxic brain inflammation, prevented abnormal behaviors in the Alzheimer’s-model mice. The intervention also addressed spontaneous behavioral changes, offering a promising avenue for future treatments.

This innovative tool could decode the origins and progression of devastating brain disorders, paving the way for earlier detection and more effective therapies.