Water-tight membranes from neuronal morphology files

Area of Science:

• Computational Neuroscience
• Biophysics
• Computer Graphics

Background:

• Neuronal morphology data, often from light microscopy or algorithmic generation, is typically represented as points and diameters for electrophysiology simulations.
• This limited geometric information is insufficient for multi-scale models involving 3D reaction-diffusion processes due to smaller space constants.
• Existing methods produce geometries unsuitable for detailed biophysical simulations requiring precise spatial relationships.

Purpose of the Study:

• To develop a novel algorithm for generating watertight 3D neuronal surface meshes from point-and-diameter data.
• To enable accurate simulations of reaction-diffusion processes within reconstructed neuronal geometries.
• To bridge the gap between electrophysiology-focused morphology descriptions and the requirements of multi-scale biophysical modeling.

Main Methods:

• Introduced the Constructive Tessellated Neuronal Geometry (CTNG) algorithm.
• Utilized constructive solid geometry principles to reconstruct plausible neuronal shapes without gaps or cul-de-sacs.
• Employed a "constructive cubes" approach to generate watertight triangular surface meshes.

Main Results:

• CTNG successfully creates watertight 3D triangular meshes from sparse point-and-diameter data.
• The algorithm establishes a crucial link between internal voxels and surface triangles for integrated simulations.
• Optimized marching cubes and distance calculations improved performance for complex neuronal structures.

Conclusions:

• The CTNG algorithm provides a robust method for reconstructing detailed neuronal geometries.
• This approach facilitates accurate multi-scale modeling, particularly for 3D reaction-diffusion simulations.
• CTNG enhances the utility of existing morphology datasets for advanced biophysical research.