Latest news and developments

Josh Bernstock, MD, PhD, MPH, a neurosurgeon-scientist and entrepreneur affiliated with Texas Children’s Hospital, Baylor College of Medicine, and MD Anderson, is one of Centile Bioscience's co-founders. As Chief Medical Officer, he guides the company's clinical direction, ensuring its technology delivers practical value in clinical care, drug development, and research.

Hannah Mamuszka, a health-economics leader and diagnostics expert, joined Centile Bioscience's Advisory Board, where she guides the company's market-access, reimbursement, and health-economics strategy as Centile's platform moves toward clinical and commercial use.

Centile Bioscience was founded to carry a body of award-winning brain science into practice. Its founding research, "Brain charts for the human lifespan," was published in Nature and won the inaugural FASEB/NIH DataWorks! Grand Prize, one of only two Grand Prize awards selected from more than 100 competing research groups nationwide.

TissUnet extends quantitative MRI analysis beyond the brain, segmenting skull bone, subcutaneous fat, and muscle from routine T1-weighted scans across all ages and pathologies, achieving an 89% acceptance rate in healthy and brain tumor cohorts.

Brain growth charts reveal dynamic extra-axial CSF trajectories in clinical pediatric MRI, enabling automated macrocephaly diagnosis and challenging fixed-measurement guidelines.

Regional brain centile scores predict cognitive performance and discriminate Alzheimer's disease from matched controls, extending to frontotemporal lobar degeneration and validating individual-level clinical applicability across neurodegenerative conditions.

The first population framework for brain asymmetry from prenatal to age 102 provides new benchmarks for neuropsychiatric and neurodegenerative research.

Next-generation 3D deep learning CT brain segmentation models outperform prior methods, validated across cognitively normal and dementia patient cohorts.

A deep learning defacing model removes facial features from pediatric brain MRI with minimal research impact, enabling privacy-preserving clinical data reuse at scale.

Fine-tuned language models classify brain MRI reports with over 99.5% accuracy, enabling automated growth chart generation matching human annotation.

Normative brain charting reveals distinct neuroanatomy in autism and ADHD, including sex differences often missed, supporting individual-level characterization.

An atlas-guided skull-stripping model validated on 21,334 scans from 18 sites generalizes across all ages, scanner types, and imaging protocols, enabling reliable brain volumetry on diverse clinical archives.

Large sample sizes and longitudinal study design are the features that most reliably produce replicable neuroimaging findings, providing a scientific foundation for population-scale brain charting.

Population brain charting links structural brain changes in psychosis to their underlying molecular biology.

Centile Bio has announced a new milestone in its ongoing effort to develop quantitative reference frameworks for understanding the human brain. By integrating large-scale neuroimaging datasets with advanced statistical modeling, the company continues to expand the scientific tools available for studying brain structure, variation, and disease progression. The work contributes to a growing body of research focused on improving how neurological changes are measured across populations and throughout the human lifespan.

Deep learning extracts brain volumetric biomarkers from routine head CT with accuracy comparable to MRI, making population-scale neurodegeneration screening viable without specialist imaging.

CT-based volumetric measures obtained through deep learning: Association with biomarkers of neurodegeneration

An invited commentary mapping the clinical and research applications of normative brain charting, from individualized medicine and early detection to trial enrichment, rare disease phenotyping, and longitudinal monitoring.

Copy Number Variant Risk Scores Associated With Cognition, Psychopathology, and Brain Structure in Youths in the Philadelphia Neurodevelopmental Cohort

Deep learning from MRI-derived labels enables automatic brain tissue classification on human brain CT