Direct Reconstruction of DC Cortical Conductivity from Large-Scale Electron Microscopy Data.

Area of Science:

• Neuroscience
• Computational Biology
• Biophysics

Background:

• Electrical conductivity in cortical gray matter is crucial for understanding brain activity and stimulation.
• Existing macroscopic conductivity values vary significantly, hindering accurate bioelectromagnetic modeling.
• The source of this variability (measurement error vs. structural differences) remains unclear.

Purpose of the Study:

• To develop a multiscale computational framework for deriving mesoscale conductivity maps of the mouse visual cortex.
• To achieve 50-µm resolution conductivity mapping using segmented electron microscopy data.
• To investigate whether observed conductivity variations are due to intrinsic structural heterogeneity.

Main Methods:

• Utilized the Minnie 65 subvolume from the MICrONS dataset, subdividing it into 1,224 cubic blocks.
• Applied quasistatic electric modeling using an iterative boundary-element fast multipole method (BEM-FMM).
• Estimated the conductivity tensor components by applying orthogonal electrode pairs to each block, assuming non-conducting membranes (DC conductivity).

Main Results:

• The computational framework successfully derived mesoscale conductivity maps at 50-µm resolution.
• Spatially averaged conductivity values aligned with previous low-resolution rat studies, validating the method.
• Mesoscale maps showed significant conductivity granularity at 50-100 µm scales and directional variations.

Conclusions:

• Mesoscale conductivity heterogeneity is likely an intrinsic structural property of the cortex.
• The findings provide a higher-resolution understanding of cortical electrical properties.
• This framework can advance bioelectromagnetic modeling and the study of brain function.