This repository contains the source code and data corresponding to PLOS 2023 publication:

Population-wise Labeling of Sulcal Graphs using Multi-graph Matching.

The graph matching methods used in this work were implemented using authors implementation and can be found in their corresponding repositories:

Kernelized graph matching (Code)

Multi-image matching via fast alternating minimization (Code)

Multi-Graph Matching via Affinity Optimization with Graduated Consistency Regularization (Code)

Solving the multi-way matching problem by permutation synchronization (Code)

---

All permutation matrices, silhouette values and consistency values are precomputed and provided in the folder Oasis_original_new_with_dummy

Dependencies:

These dependencies are mandatory for the implementation of scripts provided, please follow the instructions properly provided in the following dependencies for proper reproducibility of the scripts provided:

1. SLAM

2. VISBRAIN for visualization.

To generate a population of synthetic sulcal graphs(specify the parameters in script):

python script_generation_graphs_with_edges_permutation_updated.py

Computation of affinity matrices:

python script_generation_affinity_and_incidence_matrix.py

Visualize labelings from graph matching methods:

python process_real_data/08_script_visu_labelling_template.py

Visualize silhouettes from graph matching methods:

python process_real_data/12_script_visu_silouhette.py