Non-central chi estimation of multi-compartment models improves model selection by reducing overfitting - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

Non-central chi estimation of multi-compartment models improves model selection by reducing overfitting

Résumé

Diffusion images are known to be corrupted with a non-central chi (NCC)-distributed noise [1]. There has been a number of proposed image denoising methods that account for this particular noise distribution [1,2,3]. However, to the best of our knowledge, no study was performed to assess the influence of the noise model in the context of diffusion model estimation as was suggested in [4]. In particular, multi-compartment models [5] are an appealing class of models to describe the white matter microstructure but require the optimal number of compartments to be known a priori. Its estimation is no easy task since more complex models will always better fit the data, which is known as over-fitting. However, MCM estimation in the literature is performed assuming a Gaussian-distributed noise [5,6]. In this preliminary study, we aim at showing that using the appropriate NCC distribution for modelling the noise model reduces significantly the over-fitting, which could be helpful for unravelling model selection issues and obtaining better model parameter estimates.
Fichier principal
Vignette du fichier
2623.pdf (163.61 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

inserm-00993964 , version 1 (21-05-2014)

Identifiants

  • HAL Id : inserm-00993964 , version 1

Citer

Aymeric Stamm, Benoit Scherrer, Stefano Baraldo, Olivier Commowick, Simon K. Warfield. Non-central chi estimation of multi-compartment models improves model selection by reducing overfitting. ISMRM, May 2014, Milan, Italy. pp.2623. ⟨inserm-00993964⟩
916 Consultations
245 Téléchargements

Partager

Gmail Facebook X LinkedIn More