Explore charts

Brain views show reference anatomy for selecting phenotypes, not individual measurements. Left/right labels refer to anatomical hemispheres; visualization geometry is not used to calculate model predictions.

• Modality banners. Structural MRI shows sagittal, coronal and axial slices of the public FreeSurfer fsaverage reference. The resting-state fMRI and ASL banners link this reference anatomy to illustrative BOLD and control/label/difference signals. Colored marks and signals are schematic, not measured activation, perfusion or subject data.
• Global anatomy. FreeSurfer fsaverage MRI and surfaces, rendered with NiiVue. Surfaces are smoothed illustrations. The eTIV overlay highlights available reference brain tissue, not an intracranial-cavity segmentation.
• Cortical atlases. Local anatomical DK68 and FC System use ggseg/ggsegYeo2011 vertex labels on the ggseg3d fsaverage5 reference surface, with the same switchable 2D/3D viewer as MIND Region. FC Region uses neutral fsaverage4 reference surfaces; Desikan–Killiany lobar labels appear in the metric selector for anatomical orientation. Its 2D and 3D views share the same vertex mapping. MIND System retains its ggseg3d reference. MIND Region uses the supplied DK68, DK219, DK318, HCP-MMP1.0 (360) and Schaefer200 fsaverage annotations on the same ggseg3d fsaverage5 inflated surface. Its switchable 2D and 3D views share identical vertex labels.
• Subcortical atlas. The bilateral, 14-region aseg_subcortex atlas comes from Annie G. Bryant’s subcortex_visualization. Three.js renders surfaces reconstructed from its reference volume; the 2D views retain the original medial/lateral SVG projections, not MRI cross-sections.

Bryant (2026) · Fischl et al. (2002) · FreeSurfer license · Subcortical source license

Five atlases are available for All, Chinese and Western-centric reference populations. Ages are in years. Reference curves retain the existing DK318 prediction method and 401-point, 0–80-year grid; agreement with the original training basis still needs validation before publication. Individual deviation estimation uses the available prediction metadata and page-specific eligibility checks described in the Input guide.