Neuroscience 2026: four new biomarker studies for Alzheimer's and depression
Researchers presented blood, brain-imaging and artificial intelligence approaches at Neuroscience 2026, with one Emory team using an AI system to score movement in primate models.
By Giovanni Sapere
2 min read

Neuroscientists used the annual meeting of the Society for Neuroscience, held in November 2026, to present new ways of measuring brain and nerve disorders. The work covers early Alzheimer's-related neurodegeneration, unusually preserved brain structure in older adults, mitochondrial gene changes in depression and an artificial intelligence system built to catch movement symptoms too small for the eye. Biomarkers already guide cancer and heart care, but the nervous system has resisted the same approach.
Measuring brain regions other than the hippocampus
Meaghan Morris of Johns Hopkins University reported that measuring the size of brain structures beyond the hippocampus may help detect early Alzheimer's-related neurodegeneration and could serve as a marker of disease progression MedicalXpress. The hippocampus has long been the reference region in Alzheimer's imaging, so the finding widens the set of structures worth measuring.
A second study looked at the opposite case. Karlon H. Johnson of the University of Miami Miller School of Medicine found that some cognitively healthy older adults keep a younger-appearing brain structure even when their blood shows elevated levels of biomarkers linked to neurodegeneration MedicalXpress. That result complicates any simple reading of a blood test: a marker can be raised without matching structural loss.
“Biomarkers and outcome assessments are essential tools for improving diagnosis, prognosis and monitoring disease progression and treatment response.”
— Carol Taylor-Burds, program director, NINDS Division of Translational Research
Mitochondrial genes and major depressive disorder
In samples of brain and blood from people with major depressive disorder, Maura B. Dupont of Columbia University reported changes in mitochondrial genes. The abstract describes the finding as having potential implications for biomarker development and for treatment targets MedicalXpress.
Read also on The Patent: AI Rivals Target the Yang-Mills Problem in Physics
Mitochondria are the cell structures that produce energy, and their genes sit outside the main nuclear genome. The work connects a mood disorder to a measurable molecular signal, which is the first requirement for any diagnostic test, but the abstract does not report a validated test or a clinical trial.
An AI system that scores movement in primate models
Nisarg Desai of Emory University presented an artificial intelligence system that can detect and quantify subtle motor symptoms in nonhuman primate models of neurodevelopmental and neurodegenerative disorders MedicalXpress. The stated purpose is measurement: turning movements that escape visual observation into numbers a study can compare.
The approach is a research tool in animal models, not a diagnostic product. Nothing in the abstract describes testing in patients, regulatory clearance or commercial development.
No patent numbers, and abstracts are not peer-reviewed papers
The Society for Neuroscience release names no patent, no patent application and no licensing agreement behind any of the four lines of work. Conference abstracts are short summaries presented before full peer review, so the numbers and effect sizes behind each claim are not yet public in the form a journal paper would provide.
For institutions, the sequence is usually the reverse of the publicity: a validated biomarker or a trained AI model becomes the subject of patents and technology transfer talks once the underlying method is settled. On the evidence released so far, these four projects sit at the measurement stage, with diagnosis, protection and clinical use still ahead.
AI-assisted article



