Individual Analysis of Molecular Brain Imaging Data Through Automatic Identification of Abnormality Patterns

Abstract : We introduce a pipeline for the individual analysis of positron emission tomography (PET) data on large cohorts of patients. This pipeline consists for each individual of generating a subject-specific model of healthy PET appearance and comparing the individual's PET image to the model via a novel regularised Z-score. The resulting voxel-wise Z-score map can be interpreted as a subject-specific abnormality map that summarises the pathology's topographical distribution in the brain. We then propose a strategy to validate the abnormality maps on several PET tracers and automatically detect the underlying pathology by using the abnormality maps as features to feed a linear support vector machine (SVM)-based classifier. We applied the pipeline to a large dataset comprising 298 subjects selected from the ADNI2 database (103 cognitively normal, 105 late MCI and 90 Alzheimer's disease subjects). The high classification accuracy obtained when using the abnormality maps as features demonstrates that the proposed pipeline is able to extract for each individual the signal characteristic of dementia from both FDG and Florbetapir PET data.
Type de document :
Communication dans un congrès
Computational Methods for Molecular Imaging - [MICCAI 2017 Satellite Workshop], Sep 2017, Quebec City, Canada. 2017
Liste complète des métadonnées

Littérature citée [9 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01567343
Contributeur : Ninon Burgos <>
Soumis le : samedi 22 juillet 2017 - 19:09:34
Dernière modification le : mercredi 26 juillet 2017 - 01:11:25

Fichier

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

Identifiants

  • HAL Id : hal-01567343, version 1

Collections

| ARAMISLAB | ICM | LIB | UPMC | INRIA

Citation

Ninon Burgos, Jorge Samper-González, Anne Bertrand, Marie-Odile Habert, Sébastien Ourselin, et al.. Individual Analysis of Molecular Brain Imaging Data Through Automatic Identification of Abnormality Patterns. Computational Methods for Molecular Imaging - [MICCAI 2017 Satellite Workshop], Sep 2017, Quebec City, Canada. 2017. 〈hal-01567343〉

Partager

Métriques

Consultations de la notice

135

Téléchargements de fichiers

30