Visualizing cortical laminar architecture in the living human brain using next-generation ultra-high-gradient diffusion MRI

Characterizing cortical laminar microstructure is essential for understanding the organization of the human brain. Leveraging the next-generation Connectome MRI scanner (maximum gradient strength=500mT/m, slew rate=600 T/m/s), we characterized in vivo cortical laminar cytoarchitecture and myeloarchitecture through cortical depth-dependent analyses of soma and neurite density imaging (SANDI) metrics derived from 1 mm diffusion MRI generated using a super-resolution technique. SANDI revealed distinct laminar profiles: intra-soma signal fraction peaked at ~55% cortical depth, while the intra-neurite signal fraction increased toward deeper cortical layers, consistent with known histological patterns. The visual cortex showed higher intra-soma signal fraction than the motor cortex, particularly in deeper layers. Intra-soma signal fraction correlated positively with cortical curvature in superficial layers and negatively in deeper layers, indicating layer-specific relationships between cortical microstructure and geometry. These findings demonstrate the feasibility of noninvasive mapping of laminar architecture, offering a potential in vivo surrogate for histology and enabling future studies of cortical laminar organization using high-performance gradient MRI.