Image-encoded biological and non-biological variables may be used as shortcuts in deep learning models trained on multisite neuroimaging data

OBJECTIVE: This work investigates if deep learning (DL) models can classify originating site locations directly from magnetic resonance imaging (MRI) scans with and without correction for intensity differences. MATERIAL AND METHODS: A large database of 1880 T1-weighted MRI scans collected across 41...

Descripción completa

Detalles Bibliográficos
Autores principales: Souza, Raissa, Wilms, Matthias, Camacho, Milton, Pike, G Bruce, Camicioli, Richard, Monchi, Oury, Forkert, Nils D
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Oxford University Press 2023
Materias:
Research and Applications
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10654841/
https://www.ncbi.nlm.nih.gov/pubmed/37669158
http://dx.doi.org/10.1093/jamia/ocad171

Internet

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10654841/
https://www.ncbi.nlm.nih.gov/pubmed/37669158
http://dx.doi.org/10.1093/jamia/ocad171

Ejemplares similares