Table 1:

Advantages and disadvantages of three approaches for whole-brain cortical segmentation/parcellation. Research activities focus on improving the spatial resolution and SNR efficiency of in vivo high-resolution dMRI acquisitions (D. A. Feinberg et al., 2023; Huang et al., 2021; F. Wang et al., 2021), including MAP-MRI (Huang et al., 2021), and on developing automated processing pipelines and computational methods for 2D and 3D analysis of histological data (Schiffer et al., 2021; Schleicher et al., 2005; Wagstyl et al., 2020).

	Histological Analysis	Warping-based parcellation	Direct MAP-based segmentation
Contrast mechanism	Molecular stains	GM/WM anatomy	Water diffusion
Anatomical landmarks	Cellular and subcellular (≈1μm)	Cortical curvature (≈1mm)	Tissue architecture (≈200μm)
Spatial scale	Microscopic	Macroscopic	Mesoscopic
Field-of-view (FOV)	2D section	3D whole-brain	3D whole-brain
Processing	Slicing, Staining, Registration, Digitization, Statistical Analysis, Segmentation	Scanning, Template warping	Scanning, Distortion/motion correction, Direct clustering
Acquisition duration	Days-weeks	Minutes/hours	Hours
Cost	≈ $50,000	≈ $200	≈ $500
In vivo compatibility	N/A	Structural scan ≈5 min	Requires further improvements in resolution and SNR efficiency