Locus




                          



About the Software

The LOCUS software analyses in few minutes a 3D MR brain scan and identifies brain tissues and a large number of brain structures. An image is divided into cubes on each of which a statistical model is applied. This provides a number of local treatments that are then integrated to ensure consistency at a global level. It results a low sensitivity to artefacts. The statistical model is based on a Markovian approach which enables to capture the relations between tissues and structures, to integrate a priori anatomical knowledge and to handle local estimations and spatial correlations.

The main functionalities of the program include :

  • Tissue and structure segmentation combined with registration to a structure atlas
  • Robustness to biais field inhomogeneity and noise

A flyer in French and English, a slide, and a video are available for a short description. The software is written in C++ (50000 lines).


LOCUS has been used by the team in various neuroimaging applications where gray matter segmentation is an issue. Extensions are undergoing for lesions segmentation in brain pathologies such as Multiple Sclerosis and Stroke (see Forbes F, Doyle S, Garcia-Lorenzo D, Barillot C and Dojat M, A Weighted Multi-Sequence Markov Model For Brain Lesion Segmentation, Thirteenth International Conference on Artificial Intelligence and Statistics (AISTATS 2010), Sardinia, Italy (May 13-15), (2010) to appear).




Information / News

      A new version locus3-2.0 is available.
This version includes a graphical user interface and is available for Windows, Mac0S and Linux.



Publications

The seminal method and following refinements are described in several publications:

  • Scherrer. B.Segmentation des tissus et structures sur les IRM cerebrales. Phd Thesis, Institut National Polytechnique de Grenoble (INPG), Dec. 2008.
  • Scherrer, B; Forbes, F; Dojat, M, A Conditional Random Field approach for coupling local registration with robust tissue and structure segmentation., MICCAI , (2009) Springer Verlag, London (UK), 2009:540-48.pdf
  • Scherrer, B; Forbes, F; Garbay, C; Dojat, M, Distributed Local MRF Models for Tissue and Structure Brain Segmentation, IEEE Trans Med Imaging, (2009). pdf
  • Scherrer, B; Dojat, M; Forbes, F; Garbay, C, Agentification of Markov model-based segmentation: application to magnetic resonance brain scans, Artif Intell Med, (2009)
  • Scherrer, B; Forbes, F; Garbay, C; Dojat, M, Fully Bayesian joint model for MR brain scan tissue and structure segmentation, MICCAI , (2008)Springer-Verlag, New York City, 2008:1066-74. pdf
  • Scherrer, B; Dojat, M; Forbes, F; Garbay, C, LOCUS: local cooperative unified segmentation of MRI brain scans, MICCAI , (2007) Springer-Verlag, Berlin, 2007:13-23. pdf
  • Forbes F. and Peyrard N. Hidden Markov Random Field selection criteria based on Mean Field-like approximations. IEEE transactions on Pattern Analysis and Machine Intelligence, vol.25, no.8, August 2003. pdf


About the Team

LOCUS is the result of several years of collaborative research; first initiated with a conjoint work between the NeuroImaging team at GIN (expert in Neuroimaging data analysis) and the MAGMA team at LIG (expert in multi-agent approaches), the statistical model was strongly extended with the participation of the MISTIS team at INRIA Rhône-Alpes (expert in advanced statistical methods).


Documentation

A documentation is available on the use of LOCUS : Documentation.


Download

LOCUS is available upon request.


Contact

If you have problems, questions or remarks on the software, send an email to: LOCUS-help@lists.gforge.inria.fr





Last modification : April 2010

 

 


Powered By GForge Collaborative Development Environment