Researchers have leveraged artificial intelligence to uncover a previously unknown cause of Alzheimer’s disease, potentially revolutionizing how the condition is understood and treated. Scientists from Massachusetts General Hospital and Harvard Medical School used a machine learning model called SuStaIn (Subtype and Stage Inference) to analyze thousands of brain scans and medical records from Alzheimer’s patients.
The AI identified a new disease pathway distinct from the well-known amyloid and tau protein build-up traditionally linked to the condition. The newly discovered subtype revolves around the degeneration of a brain structure linked to memory, called the fornix, triggering a unique progression of cognitive decline.
This discovery could have major implications for diagnosis and personalized treatment, offering a path to earlier intervention based on a patient’s specific disease subtype. It also suggests that Alzheimer’s is even more biologically diverse than previously believed. The researchers hope that AI-based approaches like this will accelerate the development of customized therapies and improve clinical trial design by targeting patients more precisely.
