AI-built molecular atlas maps Alzheimer’s brain beyond amyloid plaques, pointing to overlooked metabolic shifts

Researchers at Rice University have created a label-free molecular atlas of the Alzheimer’s brain in an animal model, using laser-based imaging paired with artificial intelligence. The work aims to clarify how the disease emerges and spreads beyond what standard pathology typically captures.

The study used hyperspectral Raman imaging, an advanced form of Raman spectroscopy that reads chemical fingerprints in tissue without dyes or fluorescent tags. By scanning brain slices at high resolution, the team generated a detailed chemical map designed to reflect the brain’s native state.

What the imaging revealed

Analysis indicated that Alzheimer’s-linked chemical changes were not limited to amyloid plaques. Instead, the alterations appeared across multiple brain regions, with uneven patterns that could help explain why symptoms develop gradually and differ between individuals.

To handle the large dataset, the researchers applied both unsupervised and supervised machine learning methods. Unsupervised tools grouped tissue by molecular similarity, while supervised models helped distinguish Alzheimer’s-affected samples from controls across different regions.

Metabolic signals in key regions

Beyond protein-related pathology, the maps pointed to broader metabolic differences, including shifts in cholesterol and glycogen signals. The strongest contrasts were reported in brain regions central to memory and cognition, including the hippocampus and cortex.

The authors argue that these molecular patterns support a wider view of Alzheimer’s as a disorder involving disrupted brain structure and energy balance, not only plaque formation. They say a whole-brain, label-free approach could help surface changes that targeted assays might miss.

While the findings are based on an animal model and would need validation in human tissue, the researchers suggest the approach could eventually inform earlier detection strategies and more region-specific treatment research. The study was published in ACS Applied Materials and Interfaces with support from U.S. federal research funders.