Geometry-Based Multiscale Mammalian Brain Parcellation
exploratoryresearch program · medium · Wed Jan 01 2025 00:00:00 GMT+0000 (Coordinated Universal Time)
Generate multiscale parcellations of mammalian brain structures using geometry alone and study geometric influences on regional brain organization.
A data-driven geometric approach for producing brain parcellations across species and resolution scales, benchmarked against anatomical, functional, cellular, and molecular properties in humans, macaques, marmosets, and mice.
2025 bioRxiv preprint describing the geometry-based parcellation method and cross-species benchmarks.
The geometry-based regions were reported to be more homogeneous than nearly all benchmark parcellations across hundreds of measured brain properties, and the method generalized to species without existing parcellations.
White Matter Morphometry Normative Modeling in Psychiatric Disorders
exploratoryresearch program · medium · Wed Jan 01 2025 00:00:00 GMT+0000 (Coordinated Universal Time)
Characterize person-specific white matter volume deviations and heterogeneity across psychiatric disorders using MRI-derived normative models.
Warped Bayesian linear regression normative models applied to T1-weighted MRI data from 1,294 psychiatric cases and 1,465 matched controls across 25 scan sites; deviations analyzed across voxels, interregional connections, regions, and networks.
Peer-reviewed publication in 2025 reporting multiscale white matter morphometry heterogeneity across six psychiatric disorders.
White matter volume deviations were highly heterogeneous, with the same voxel affected in fewer than 8% of individuals within a diagnosis; autism and schizophrenia showed aggregation of negative deviations into common tracts, regions, and networks in subsets of individuals.
Whole-Brain Multimodal MRI Discrimination of Parkinson's Disease and Multiple System Atrophy
exploratoryresearch program · medium · Tue Jan 01 2019 00:00:00 GMT+0000 (Coordinated Universal Time)
Discriminate Parkinson's disease, multiple system atrophy, and healthy controls using interpretable multimodal MRI-derived features.
Whole-brain multimodal MRI pipeline combining feature selection and reduction across grey matter volume, fractional anisotropy, mean diffusivity, iron deposition, resting-state brain activity, and connectivity.
2019 NeuroImage: Clinical publication reporting classification performance for PD, MSA, and healthy controls.
Reported best accuracies were 0.78 for Parkinson's disease versus healthy controls, 0.94 for multiple system atrophy versus healthy controls, and 0.88 for Parkinson's disease versus multiple system atrophy.
BrainKey MRI Brain Health Platform
undisclosedplatform · high
Turn complex brain MRI data into clear visualizations and AI-derived brain health insights, including biomarkers and anatomy-level information, for educational and informational use rather than diagnosis.
AI analysis of brain MRI scans to examine biomarkers and generate intuitive visualizations of brain anatomy.
BrainKey publicly presents the platform on its website and lists a publications page.
