Table of Contents
Centile Bioscience converts routine clinical brain imaging into structured, quantitative, population-referenced biological measurements, transforming imaging from a qualitative diagnostic tool into scalable measurement infrastructure for precision neurology. A peer-reviewed study, co-authored by Jakob Seidlitz, PhD, Joshua D. Bernstock, Richard A.I. Bethlehem, PhD, Simon R. White, PhD, and Edward T. Bullmore, PhD, all part of Centile Bio's founding research team, validated that approach in one of the hardest settings: detecting disease-related brain change in a single patient.
The study, "Generalizable MRI normative modelling to detect age-inappropriate neurodegeneration," published in Alzheimer's Research & Therapy, tested whether that approach could generalize from a population reference to the individual scans clinicians and researchers actually work with. Researchers built regional brain charts of cortical thickness and hippocampal and amygdala volumes, referenced against normative data from 56,173 participants across the lifespan. Each patient's regional measurements were then calculated as a centile score.
The centile scores distinguished 351 Alzheimer's disease patients, confirmed by neuropathological evidence, from matched cognitively normal controls. They also predicted cognitive performance and tau-PET tracer uptake, tying the structural MRI measurement to independent markers of disease. Researchers confirmed the finding in a second, independent cohort of 39 Alzheimer's dementia patients and 71 with mild cognitive impairment, then extended the approach to 113 cases of frontotemporal lobar degeneration, recovering atrophy patterns that differed by clinical phenotype.
These results support population-referenced measurement as a source of endpoints for patient stratification, therapeutic response, and longitudinal monitoring, across two distinct diseases and multiple independent datasets. This demonstration of measurement generalization provides the type of evidence needed for pharma and biotech teams to implement this analysis in Alzheimer's disease or frontotemporal dementia trials.
This validation builds directly on the same normative-modelling method behind Centile Bioscience's flagship reference, the 2022 Nature study "Brain charts for the human lifespan." These results extend that foundation from a population reference into evidence clinicians, researchers, and drug developers can act on.
This is an exciting set of results validating a core thesis of the CentileOS platform: quantitative benchmarking of imaging phenotypes allows for precision mapping of individual atrophy patterns, concordant with established disease processes.
The demonstrated generalizability across two neurodegenerative diseases and independent patient groups, tracking with cognition and molecular pathology, is the evidence hospitals, researchers, and drug developers need before they build a clinical decision or a trial endpoint on Centile's platform.
— Jakob Seidlitz, PhD, co-founder & CEO, Centile Bioscience
Partner with us to bring precision quantification into your imaging pipeline or trial design.
