NeuroPath2Path

NeuroPath2Path analyzes neuronal morphology by encoding neurons as graphs and comparing sets of rooted soma-to-dendrite paths to quantify morphological differences for neuromorphology analysis and neuron classification.

Key Features:

• Graph-Theoretic Morphology Analysis: Encodes complex neuronal structures as graphs to enable high-throughput informatic analysis and address complex subgraph matching challenges.
• Rooted Path Decomposition: Decomposes neurons into rooted paths from the soma to dendrites and extracts morphological features from each constituent path.
• Elastic Morphing Framework: Uses the square root velocity function to perform elastic deformations and compute continuous morphs between sets of rooted paths while minimizing transformation cost.
• Path Correspondence Establishment: Establishes correspondence between paths of two neurons using a modified Munkres algorithm to align and compare rooted-path sets.

Scientific Applications:

• Neuronal Classification: Quantifies shape-based and graph-related features, including overall size, global shape, local branch patterns, and cell-specific biophysical properties, to support classification of neuron cell types.
• Developmental Neuroscience: Compares rooted-path morphology across developmental stages to investigate how neuronal structure matures and reorganizes.
• Disease Modeling: Detects and quantifies deviations in neuronal morphology to aid modeling and analysis of neurological disease-related structural changes.

Methodology:

Encode neurons as graphs, decompose graphs into rooted soma-to-dendrite paths and extract path-level morphological features, apply an elastic deformation framework using the square root velocity function to compute continuous morphs and minimize morphing cost between path sets, and establish path correspondence with a modified Munkres algorithm while addressing complex subgraph matching.