IMAGeS team: IMages, leArning, Geometry and Statistics

Difference between revisions of "TIBM: Segmentation"

From IMAGeS team: IMages, leArning, Geometry and Statistics
Jump to navigation Jump to search
Line 11: Line 11:
 
* Specific and original algorithms have also been developed in the project [http://icube-miv.unistra.fr/fr/index.php/ERC_FBrain FBrain ERC] for segmenting brain tissues from the fetal MRI [https : //icube-publis.unistra.fr/2-RHS11 RHS11-2], [2-https://icube-publis.unistra.fr/2-CPHS11 CPHS11], [https: //icube-publis.unistra .com / 2-PNRK11 2-PNRK11].
 
* Specific and original algorithms have also been developed in the project [http://icube-miv.unistra.fr/fr/index.php/ERC_FBrain FBrain ERC] for segmenting brain tissues from the fetal MRI [https : //icube-publis.unistra.fr/2-RHS11 RHS11-2], [2-https://icube-publis.unistra.fr/2-CPHS11 CPHS11], [https: //icube-publis.unistra .com / 2-PNRK11 2-PNRK11].
 
* Finally, Markov approaches have been implemented for the automatic identification of multiple sclerosis lesions in multimodal  [https://icube-publis.unistra.fr/2-BCA14-BCA14 2] and temporal [https://icube-publis.unistra.fr/4-LCA14 4-LCA14] MRI sequences.
 
* Finally, Markov approaches have been implemented for the automatic identification of multiple sclerosis lesions in multimodal  [https://icube-publis.unistra.fr/2-BCA14-BCA14 2] and temporal [https://icube-publis.unistra.fr/4-LCA14 4-LCA14] MRI sequences.
−
* Besides MRI, other contributions, based on a representation in super-voxels of the image, image segmentation involve CT scan, especially to delineate the vertebrae [https: //icube-publis.unistra .com / 4-CRMC15 4-CRMC15] or estimate the hepatic tumor necrosis rate [https://icube-publis.unistra.fr/4-CNRH16 4-CNRH16], [https: //icube-publis.unistra .com / 4-CRNH15 4-CRNH15].
+
* Besides MRI, other contributions based on super-voxel decomposition involve CT scan images, especially to delineate vertebrae [https: //icube-publis.unistra .com / 4-CRMC15 4-CRMC15] or estimate the hepatic tumor necrosis rate [https://icube-publis.unistra.fr/4-CNRH16 4-CNRH16], [https://icube-publis.unistra.com/4-CRNH15 4-CRNH15].
−
* A segmentation method of high-resolution data based on histopathological texture descriptors and a supervised classification has also been proposed [4-https://icube-publis.unistra.fr/4-ANFF14 ANFF14].
+
* A method combining texture descriptors and a supervised classification scheme has also been proposed for segmenting high-resolution histopathological data [4-https://icube-publis.unistra.fr/4-ANFF14 4-ANFF14].
  
−
 
−
* Une approche de segmentation multi-atlas basée sur de la fusion par patch a été proposée pour parcelliser le cortex ainsi que les structures profondes du cerveau [https://icube-publis.unistra.fr/2-RHS11 2-RHS11].
 
−
* Dans le cadre de l’ [http://icube-vivabrain.unistra.fr/index.php/Main_Page ANR Vivabrain], une méthode basée sur l’utilisation d’arbres de composantes (component-trees) a été mise en œuvre pour la segmentation du réseau vasculaire cérébral [https://icube-publis.unistra.fr/2-DTNT13 2-DTNT13].
 
−
* Des algorithmes spécifiques et originaux ont par ailleurs été développés dans le cadre du projet [http://icube-miv.unistra.fr/fr/index.php/ERC_FBrain ERC FBrain] pour la segmentation des tissus cérébraux en IRM fœtale [https://icube-publis.unistra.fr/2-RHS11 2-RHS11], [https://icube-publis.unistra.fr/2-CPHS11 2-CPHS11], [https://icube-publis.unistra.fr/2-PNRK11 2-PNRK11].
 
−
* Enfin, des approches markoviennes ont été mises en œuvre pour l’identification automatique de lésions de sclérose en plaques dans des séquences d’IRM multimodale [https://icube-publis.unistra.fr/2-BCA14 2-BCA14] et temporelle [https://icube-publis.unistra.fr/4-LCA14 4-LCA14].
 
−
* Outre l’IRM cérébrale, d’autres contributions, basées sur une représentation en super-voxels de l’image, concernent la segmentation d’image scanner X, notamment afin de délimiter les vertèbres [https://icube-publis.unistra.fr/4-CRMC15 4-CRMC15] ou estimer le taux de nécrose de tumeurs hépatiques [https://icube-publis.unistra.fr/4-CNRH16 4-CNRH16], [https://icube-publis.unistra.fr/4-CRNH15 4-CRNH15].
 
−
* Une méthode de segmentation de données histopathologiques haute résolution reposant sur des descripteurs de texture et une classification supervisée a par ailleurs été proposée [https://icube-publis.unistra.fr/4-ANFF14 4-ANFF14].
 
 
</td>
 
</td>
 
<td>
 
<td>

Revision as of 15:55, 7 October 2016


This research topic gathers all contributions related to the automatic extraction of anatomical or functional structures from biomedical images.


  • A multi-atlas segmentation framework based on patch fusion has been proposed for parcellating both the cortex and the deep brain structures 2-RHS11.
  • In the context of the Vivabrain ANR, a method based on component-trees has been set upt to segment the cerebral vascular tree 2-DTNT13.
  • Specific and original algorithms have also been developed in the project FBrain ERC for segmenting brain tissues from the fetal MRI [https : //icube-publis.unistra.fr/2-RHS11 RHS11-2], [2-https://icube-publis.unistra.fr/2-CPHS11 CPHS11], [https: //icube-publis.unistra .com / 2-PNRK11 2-PNRK11].
  • Finally, Markov approaches have been implemented for the automatic identification of multiple sclerosis lesions in multimodal 2 and temporal 4-LCA14 MRI sequences.
  • Besides MRI, other contributions based on super-voxel decomposition involve CT scan images, especially to delineate vertebrae [https: //icube-publis.unistra .com / 4-CRMC15 4-CRMC15] or estimate the hepatic tumor necrosis rate 4-CNRH16, 4-CRNH15.
  • A method combining texture descriptors and a supervised classification scheme has also been proposed for segmenting high-resolution histopathological data [4-https://icube-publis.unistra.fr/4-ANFF14 4-ANFF14].
thumb

PhD thesis

→ Back to Biomedical image processing page

Retrieved from "https://images.icube.unistra.fr/en/index.php?title=TIBM:_Segmentation&oldid=759"