Researchers at Washington University School of Medicine in St. Louis have developed an AI-driven blood test capable of distinguishing between four major neurodegenerative diseases with 92.3% accuracy. The study, published in Alzheimer's & Dementia, details a classifier designed to identify Alzheimer's disease, Parkinson's disease, frontotemporal dementia, and dementia with Lewy bodies. Unlike current diagnostic tools that often force a single diagnosis, this new technology can detect mixed pathologies where multiple diseases occur simultaneously, a common but clinically challenging scenario that complicates patient treatment. Dr. Carlos Cruchaga, the study's senior author, emphasized that current methods frequently label patients with one condition while overlooking a complex mixture of brain injuries. The goal was to create a test that provides a comprehensive picture of all major neurodegenerative processes occurring in a person's brain, enabling precision diagnosis and personalized treatment strategies. To build the test, the team analyzed 15 specific proteins found in the blood that reflect neurodegenerative pathology, including markers for Alzheimer's, synapse damage, and inflammation. They trained and tested the AI model using data from over 3,200 individuals, covering patients with clinical diagnoses of the four target diseases as well as cognitively normal controls. The results were then verified against a separate group of 225 individuals whose brains were examined at autopsy after their clinical evaluations. The classifier's outputs aligned closely with the actual pathological burden found in brain tissue. The tool proved particularly valuable in ambiguous cases. For patients with mild cognitive impairment or unclear neurological diagnoses, the model's predictions regarding Alzheimer's pathology closely matched the amyloid plaque counts found during autopsy. Additionally, the test identified Alzheimer-like biological changes in patients originally diagnosed with Parkinson's disease who later developed dementia, highlighting its ability to detect mixed pathology that clinical assessments alone might miss. While the results are promising, the test is not yet ready for routine clinical use. Dr. Cruchaga noted that further validation in larger and more diverse populations is necessary to confirm generalizability. Prospective studies tracking patients over time will also be required to assess how well the tool predicts disease progression and guides therapeutic decisions. If validated, the blood-based classifier offers broad applications for both research and clinical practice. In research, it could help identify suitable candidates for clinical trials targeting specific disease pathways and facilitate large-scale population studies that are currently impractical due to the high costs and invasiveness of brain scans or spinal taps. In clinical settings, the tool could assist physicians in determining which patients require further follow-up, which specialists to consult, and which preventive strategies might be most effective.