Towards a Faster Randomized Parcellation Based Inference

Abstract : In neuroimaging, multi-subject statistical analysis is an essential step, as it makes it possible to draw conclusions for the population under study. However, the lack of power in neuroimaging studies combined with the lack of stability and sensitivity of voxel-based methods may lead to non-reproducible results. A method designed to tackle this problem is Randomized Parcellation-Based Inference (RPBI), which has shown good empirical performance. Nevertheless, the use of an agglomerative clustering algorithm proposed in the initial RPBI formulation to build the parcellations entails a large computation cost. In this paper, we explore two strategies to speedup RPBI: Firstly, we use a fast clustering algorithm called Recursive Nearest Agglomeration (ReNA), to find the parcellations. Secondly, we consider the aggregation of p-values over multiple parcellations to avoid a permutation test. We evaluate their the computation time, as well as their recovery performance. As a main conclusion, we advocate the use of (permuted) RPBI with ReNA, as it yields very fast models, while keeping the performance of slower methods.
Type de document :
Communication dans un congrès
PRNI 2017 - 7th International Workshop on Pattern Recognition in NeuroImaging, Jun 2017, Toronto, Canada. 2017
Liste complète des métadonnées

Littérature citée [15 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01552237
Contributeur : Andres Hoyos Idrobo <>
Soumis le : lundi 3 juillet 2017 - 10:17:33
Dernière modification le : lundi 4 juin 2018 - 15:42:02
Document(s) archivé(s) le : jeudi 14 décembre 2017 - 18:28:46

Fichier

paper.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01552237, version 1

Citation

Andrés Hoyos-Idrobo, Gaël Varoquaux, Bertrand Thirion. Towards a Faster Randomized Parcellation Based Inference. PRNI 2017 - 7th International Workshop on Pattern Recognition in NeuroImaging, Jun 2017, Toronto, Canada. 2017. 〈hal-01552237〉

Partager

Métriques

Consultations de la notice

393

Téléchargements de fichiers

100