Communication Dans Un Congrès Année : 2025

Gradual modality dropout for segmenting ischemic stroke lesions in an unseen center with missing modalities

Résumé

In clinical practice, imaging modalities may not always be available for every patient due to scheduling, cost, or patient-specific constraints. Additionally, multi-center imaging studies often face inconsistencies in protocols, machine settings, and artifacts, compromising data quality. We propose a 3D U-Net model for ischemic lesion segmentation using a novel training technique, gradual modality dropout, which progressively deactivates imaging modalities during training. This approach ensures robust performances when all modalities are present and improves segmentation accuracy in scenarios where one or more modalities are missing in unfamiliar contexts. The model demonstrates adaptability and reliability when trained on MRI scans of stroke patients across different phases (hyper-acute,sub-acute, acute, and post-treatment) and various hospital settings. Code available here: https://github.com/sofiavarib/Gradual-modality-dropout

Fichier principal
Vignette du fichier
22_Gradual_modality_dropout.pdf (319.17 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05185389 , version 1 (25-07-2025)

Licence

Identifiants

  • HAL Id : hal-05185389 , version 1

Citer

Sofia Vargas Ibarra, Vincent Vigneron, Hichem Maaref, Sonia Garcia-Salicetti, Faria, Andreia. Gradual modality dropout for segmenting ischemic stroke lesions in an unseen center with missing modalities. 8th International Conference on Medical Imaging with Deep Learning (MIDL 2025), Jul 2025, Salt Lake City, United States. ⟨hal-05185389⟩
842 Consultations
269 Téléchargements

Partager

  • More