Solving the Cross-Subject Parcel Matching Problem: Comparing Four Methods Using Extrinsic Connectivity - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2018

Solving the Cross-Subject Parcel Matching Problem: Comparing Four Methods Using Extrinsic Connectivity

Résumé

Matching parcels across different subjects is an open problem in neuroscience. Even when produced by the same technique, parcellations tend to differ in the number, shape, and spatial localization of parcels across subject. To the best of our knowledge, no technique has been able to tackle this problem. In this work, we propose and compare four different methods to match parcels across subjects based on their extrinsic connectivity fingerprint. We test their performance by matching parcels of the Desikan atlas [4] across subjects, as its parcels have been shown to have consistent connectivity profiles across subjects [6]. Then, we repeat the experiment for a pre-existing groupwise parcellation based on structural connectivity [5], which has been show to be consistent across groups of subjects. The outline of our work can be seen in Figure 1. Our results show that our methods based on Optimal Transport (OT) and Kullback Leibler (KL) divergence achieve more than 99% of correct matches.
Fichier principal
Vignette du fichier
ohbm18_gayraud.pdf (425.91 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01737366 , version 1 (05-04-2018)
hal-01737366 , version 2 (09-04-2018)

Identifiants

  • HAL Id : hal-01737366 , version 2

Citer

Nathalie T H Gayraud, Guillermo Gallardo, Maureen Clerc, Demian Wassermann. Solving the Cross-Subject Parcel Matching Problem: Comparing Four Methods Using Extrinsic Connectivity. OHBM 2018, Jun 2018, Singapore, Singapore. ⟨hal-01737366v2⟩
647 Consultations
172 Téléchargements

Partager

Gmail Facebook X LinkedIn More