SPHARM-PDM

SPHARM-PDM performs spherical harmonic encoding and point distribution model generation to quantify and compare anatomical surface morphology in neuroimaging.

Key Features:

• Spherical Harmonic Representation: Encodes surface locations using spherical harmonics (SPHARM) for compact mathematical representation of complex shapes.
• Triangulated Surface Sampling: Samples SPHARM descriptions to produce triangulated surface point distribution models (SPHARM-PDM).
• Area-Preserving Optimization: Maps surface meshes onto the unit sphere using an area-preserving optimization based on heat-equation-based methods to minimize distortion.
• Conformal Flattening Initialization: Provides a bijective, angle-preserving conformal flattening initialization as an alternative to heat-equation mappings to reduce mapping distortion for complex shapes.
• Integration with Slicer Execution Model and SALT: Integrates into a command-line workflow via the Slicer Execution Model for incorporation into the SALT shape analysis toolbox.
• Improved Quantitative Measures: Produces improved quantitative measures of shape for complex structures compared to previous mapping methodologies.

Scientific Applications:

• Neuroimaging Research: Analyzes brain structures such as the hippocampus for detailed morphological assessment.
• Pathology Detection: Compares anatomical shapes to detect morphological differences associated with pathology in brain structures.

Methodology:

Surface meshes from brain segmentations are mapped onto the unit sphere using an area-preserving, heat-equation-based optimization, encoded with spherical harmonics (SPHARM), sampled to produce triangulated SPHARM-PDM point distribution models, and can be initialized with a bijective conformal (angle-preserving) flattening to reduce mapping distortion.