New cell-scale method reconstructs whole-brain fiber tracts from routine histology

In a study published in Nature Methods on November 3, researchers from Shenzhen Institutes of Advanced Technology of the Chinese Academy of Sciences (CAS) and the Center for Excellence in Brain Science and Intelligence Technology of CAS developed a novel method termed cytoarchitecture-based link estimation (CABLE) that infers axonal pathways from cytoarchitectonic orientation to reconstruct three-dimensional (3D) whole-brain fiber tracts at cellular scale in primate and human tissues.

Mapping how fiber bundles connect brain regions is essential for understanding brain functions and mechanisms of neurological and psychiatric diseases. Existing neuroimaging imaging tools involve a trade-off between whole-brain coverage and microscopic detail. Diffusion magnetic resonance imaging (dMRI) has relatively coarse spatial and angular resolution. Viral tracing lacks whole-brain scalability. 

CABLE recovers complex crossings and sharp turns from routine Nissl or DAPI stains and agrees with gold-standard viral tracing. Although routine histology visualizes cell bodies rather than axons, local cell shapes show non-random orientation that aligns with nearby fiber bundles. CABLE leverages this anisotropy to infer tract directions from standard stains. 

Built on the VISoR high-throughput volumetric imaging platform, the CABLE pipeline prepares 300-µm serial sections of whole brains, clears and stains tissue with DAPI or Nissl, acquires volumetric data and stitches it. It then computes orientation distribution functions for 3D tractography without genetic labels or viruses. This pipeline is cross-species and platform-agnostic.

In macaque and marmoset brains, CABLE reconstructed densely packed, intersecting, and sharply bending fibers, and validated the trajectories against viral-tracer datasets. In neonatal hypoxic-ischemic encephalopathy samples, it detected abnormal corpus callosum fibers, suggesting diagnostic value for neuropathology.

CABLE outperforms ex vivo diffusion MRI in resolving complex geometries, while avoiding the low throughput of viral tracing. This study provides a practical route for high-precision, high-throughput connectome mapping with potential applications in psychiatry and epilepsy.