New MRI-based AI tool predicts Alzheimer risk decades early

Researchers from the University of Texas Health Science Center, UTHealth Houston, the Keck School of Medicine of USC, and the University of Maryland have created a Regional Vulnerability Index (RVI‑AD) that uses routine MRI scans, large neuroimaging databases and artificial‑intelligence analysis to identify brain‑structure patterns characteristic of Alzheimer’s disease.

The RVI compares a person’s brain morphology—including thickness, volume and regional structure—to patterns observed in diagnosed patients. In a study published in Molecular Psychiatry, the index predicted elevated risk with about 70 % accuracy for individuals with mild cognitive impairment and detected early changes in people as young as 30. The approach relies on widely available, non‑invasive MRI, offering a cheaper and more accessible alternative to PET scans or spinal‑fluid tests. Validation in larger clinical settings is still required before routine screening can be implemented.