Comparison of Brain Age With Standard MRI Assessment for Dementia Risk Stratification in Subjective and Mild Cognitive Impairment
Neurology, 2026
Abstract
Introduction
Identifying patients at risk of progressing to dementia at the earliest possible stage is essential for timely clinical intervention. Brain-predicted age difference (brain-PAD) — the gap between a computer-estimated brain age derived from structural MRI and a patient's chronological age — has emerged as a potential biomarker linked to clinical decline. This study examines whether brain-PAD, as computed by Combinostics' cNeuro cMRI platform, carries prognostic value for dementia progression in patients presenting with either subjective cognitive decline (SCD) or mild cognitive impairment (MCI), and whether it adds meaningful information beyond conventional visual rating scales already used in clinical practice.
Methods
The researchers conducted a retrospective cohort study drawing on patients with SCD and MCI from the Amsterdam Dementia Cohort and the SCIENCe project, with follow-up periods of up to ten years. Brain-PAD scores were calculated from baseline MRI scans using two pretrained computational methods. Cox proportional hazards models were used to evaluate the relationship between brain-PAD and the risk of progressing to dementia, with adjustments made for age, sex, Mini Mental State Examination scores, and visual rating scales. Interactions between brain-PAD and baseline diagnosis were explored to determine whether the metric's value differed by disease stage, and its added benefit over amyloid biomarker status was assessed in a relevant patient subset.
Results
MRI-derived brain-PAD was a significant predictor of progression to dementia across memory clinic patients with SCD or MCI. Notably, in patients with SCD — the earliest and most challenging group to risk-stratify — brain-PAD delivered additional predictive power over and above that provided by standard visual rating scales alone.
Conclusion
AI-generated brain age metrics derived from routine structural MRI scans can meaningfully enhance dementia risk stratification workflows, particularly for patients at the subjective cognitive decline stage. For radiology and neurology departments managing memory clinic caseloads, integrating tools such as cNeuro cMRI could support more efficient triage, earlier intervention planning, and better allocation of diagnostic and care resources without requiring additional imaging acquisitions.
