Fuzzy Tracts

Streamlines selection with unsupervised clustering and fuzzy functional varifolds.

Polylines segmentation combining spatial fuzzy logic and clustering techniques. Streamlines coherence evaluated using spatial fuzzy sets based on positional a priori. Similarity matrix between streamlines built using the following varifold formulation:

DBSCAN for streamlines clustering and fibers prototypes selection. Prototypes selected/discarded based on their fuzzy score.

Method developed for the recognition of neural connections from MRI scans. Data from Human Connectome Project

If you use this work please cite:

White Matter Multi-resolution Segmentation using Fuzzy Theory Alessandro Delmonte, Corentin Mercier, Johan Pallud, Isabelle Bloch and Pietro Gori. ISBI 2019 - IEEE International Symposium on Biomedical Imaging, April 2019, Venice, IT.

Segmentation of White Matter Tractograms Using Fuzzy Spatial Relations Alessandro Delmonte, Isabelle Bloch, Dominique Hasboun, Corentin Mercier, Johan Pallud and Pietro Gori. OHBM 2018 - Organization for Human Brain Mapping, June 2018, Singapore, SG.

Installation and Usage.

The project may use external tools. Check WMQL and MRTrix3 guides for installation.

How to run:

$ cd FuzzyTracts
$ pip install -r requirements.txt
$ vim config.json
$ python fuzzy_tracts.py <input_tractogram>

Run the provided example with

$ python fuzzy_tracts.py test_data/input_test.vtk -fa -s -r

Contacts

For any inquiries please contact: Alessandro Delmonte @ [email protected]

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details