as-PSOCT: Volumetric microscopic imaging of human brain architecture and connectivity

Abstract

Polarization sensitive optical coherence tomography (PSOCT) with serial sectioning has enabled the investigation of 3D structures in mouse and human brain tissue samples. By using intrinsic optical properties of back-scattering and birefringence, PSOCT reliably images cytoarchitecture, myeloarchitecture and fiber orientations. In this study, we developed a fully automatic serial sectioning polarization sensitive optical coherence tomography (as-PSOCT) system to enable volumetric reconstruction of human brain samples with unprecedented sample size and resolution. The 3.5 μm in-plane resolution and 50 μm through-plane voxel size allow inspection of cortical layers that are a single-cell in width, as well as small crossing fibers. We show the abilities of as-PSOCT in quantifying layer thicknesses of the cerebellar cortex and creating microscopic tractography of intricate fiber networks in the subcortical nuclei and internal capsule regions, all based on volumetric reconstructions. as-PSOCT provides a viable tool for studying quantitative cytoarchitecture and myeloarchitecture and mapping connectivity with microscopic resolution in the human brain.

Fig. 1

System schematic of as-PSOCT. The dashed block shows the home built spectral domain PSOCT. C, collimator; CM, concave mirror; DC, dispersion compensation block; FC, fiber coupler; G, grating; LSC, line scan camera; P, polarizer; PC, polarization controller; PMC, polarization-maintaining fiber coupler; PS, polarization splitter; QWP, quarter-wave plate; SLD, superluminescent diode.

Fig. 2

The workflow pipeline of as-PSOCT.

Fig. 3

as-PSOCT imaging of a 15 cm² parasagittal section of the human cerebellum showing cytoarchitecture and in-plane fiber orientations by retardance, optic axis orientation, respectively, and validated by Nissl stain and Gallyas stain. A) The yellow arrowheads indicate the location of the dentate nucleus which forms a thin layer. The orientation map was color coded according to the color wheel, with the intensity weighted by retardance. B) The cortical layers and the fiber tracts are shown in a single folium by MIP, AIP, retardance, optic axis orientation, Nissl stain and Gallyas stain. The red arrowheads indicate the Purkinje cell layer. Labels: f, fiber tracts; g, granular layer; m, molecular layer. Scale bars: A, 8 mm; B, 1 mm.

Fig. 4

Retardance images of a block of cerebellar lobule (16 × 10 × 1.8 mm³). A) The folded cerebellar cortex is shown on orthogonal viewing planes (xy, coronal; xz, axial; yz, sagittal). The locations of the axial and sagittal planes are indicated by the dashed lines on xy-plane. The panels on the right demonstrate the Purkinje cell layer (white band indicated by the arrowhead), viewed on cross-sections of the stacked slices. The locations of the Purkinje cells are indicated by the crosses with corresponding colors on the xy-plane. Scale bars: left, 2.5 mm; right, 0.6 mm. B) Volume rendering of segmented molecular layer, granular layer and white matter for the cerebellar lobule.

Fig. 5

Thickness estimation for molecular and granular layers of cerebellar cortex. A) Voxel-based thickness map shows on a slice located in the middle of the block. The molecular layer (mol) is color coded in jet, and the granular layer (gra) in hot. The white matter is shown in white as an anatomical reference. Scale bar: 2.5 mm. B) Comparison of thickness distribution between molecular layer and granular layer in boxplot.

Fig. 6

Comparison of 3D and 2D thickness estimation for the molecular layer and granular layer of the cerebellar cortex. 3D estimation was conducted on the volumetric data, while 2D estimation was performed on slices either parallel (2D_view1) or orthogonal (2D_view2 and 2D_view3) to the physical sectioning plane.

Fig. 7

Optic axis orientation map (A-C) and 2D tractography (D-F) in the cerebellum. The orientation is encoded according to the color wheel, and the brightness is weighted by retardance. The label of * indicates the deep nucleus where retardance is low. The differed colors of the orientation indicate fiber splitting and merging (B and white traces drawn in C), which are clearly revealed by tractography (D-F). D) Tracts showing inter-folium connectivity patterns. E) Split paths within a fiber bundle. F) The pathways to the cerebellar deep nuclei. The white arrows in D-F indicate the position of seeds through which fiber tracts are created. Scale bars: A, 2 mm; B, 800 µm; C, 300 µm.

Fig. 8

Consecutive slices of fiber orientation maps in a cerebellar block (10 × 10 × 3.6 mm³). A) The maps shown on parasagittal sections are separated by 180 µm and labeled with slice numbers, from medial to lateral. Each slice represents a 60 µm thickness. The orientation is encoded according to the color wheel, and the brightness is weighted by retardance. Scale bar: 2.5 mm. B) 2D tractography reveals directional changes of fiber trajectory across slices. The origin of Z = 0 is the first slice. The white arrows indicate the location of seeds through which fiber tracts are created.

Fig. 9

Volumetric retardance image (A) and optic axis orientation maps (B, C) delineate the fiber networks in the thalamus and internal capsule. A) Volumetric reconstruction revealing 3D organization of fiber tracts in thalamus and internal capsule regions. Sections 1 and 2 show the fiber tracts of the external medullary lamina (eml) and the internal capsule (ic), respectively, on the yz-plane representing cross sections of stacked slices. B) The orientation is encoded according to the color wheel, and the brightness is weighted by AIP. C) Zoom-in of the optic axis orientation in consecutive slices revealing fiber-crossing patterns. The arrowheads indicate the locations of crossing. The origin of Z = 0 is at the first slice. The map uses the same color and intensity coding as in (B). The location of those panels is indicated by the white ROI on B. Scale bars: A, 2 mm; B, 2.5 mm; C, 1 mm. S: superior; I: inferior; L: left; R: right; A: anterior; P: posterior.

Fig. 10

as-PSOCT 3D tractography with 30 µm resolution (A) compared with registered dMRI tractography at 750 µm resolution (B). C) The location where tractography is applied in the brain is indicated on the anatomical MRI image. D) i – iv demonstrate various fiber trajectories revealed by as-PSOCT that are not resolvable with dMRI. The tracts are created by the seeds (spheres i, ii-1, ii-2, iii, and iv) selected on (A). v shows the zoom-in view of fiber crossing on ROI-i.