AI Tool Aims to Match Alzheimer's Patients to the Right Treatment

Alzheimer's disease affects millions of older adults, yet even as new therapies emerge, clinicians lack reliable tools to predict which patients will benefit from which treatments. A new phase of collaboration between NeuroXT, a South Korean biotech, and Beth Israel Deaconess Medical Center in Boston aims to close that gap using artificial intelligence and medical imaging.

The core technology involves training AI to read standard MRI scans in ways that extract insights previously requiring PET imaging — a more expensive and less accessible modality. PET scans can reveal Alzheimer's-related biomarkers deep in brain tissue, but their cost and limited availability make widespread use impractical. If AI can reliably replicate those insights from MRIs, advanced cognitive assessment becomes far more scalable.

The expanded agreement moves beyond diagnosis into treatment optimization. Researchers will analyze longitudinal data from BIDMC patients who have received Alzheimer's therapies for over a year, looking for imaging patterns that correlate with treatment response. The central question shifts from "does this person have Alzheimer's?" to "which treatment is most likely to help this particular person?"

This mirrors a transformation already underway in oncology, where personalized treatment strategies have become standard practice. Alzheimer's medicine appears to be entering a similar phase — not just developing new drugs, but learning to deploy them more precisely. The timing matters because several anti-amyloid therapies have recently entered clinical practice, making patient selection increasingly consequential.

From a longevity perspective, cognitive preservation is as critical as physical healthspan. The ability to maintain memory, independence, and quality of life into old age is a core longevity goal. AI-driven precision in Alzheimer's care could meaningfully extend functional cognitive years. Caveats remain: this is an early-stage research collaboration, and clinical validation across diverse populations will be essential before these tools influence standard care.