Communication Dans Un Congrès Année : 2024

A recurrent network for segmenting the thrombus on brain MRI in patients with hyper-acute ischemic stroke

Résumé

In the stroke workflow, timely decision-making is crucial. Identifying, localizing, and measuring occlusive arterial thrombi during initial imaging is a critical step that triggers the choice of therapeutic treatment for optimizing vascular re-canalization. We present a recurrent model that segments the thrombus in patients suffering from a hyper-acute stroke. A cross-attention module is defined to merge the diffusion and susceptibility-weighted modalities available in magnetic resonance imaging (MRI), which are fed to a modified version of convolutional long-short-term memory (CLSTM). It detects almost all the thrombi with a Dice higher than 0.6. The lesion segmentation prediction reduces the false positives to almost zero and the performance is comparable between distal and proximal occlusions.
Fichier non déposé

Dates et versions

hal-04939296 , version 1 (10-02-2025)

Identifiants

  • HAL Id : hal-04939296 , version 1

Citer

Sofia Vargas Ibarra, Vincent Vigneron, Hichem Maaref, Jonathan Kobold, Sonia Garcia-Salicetti, et al.. A recurrent network for segmenting the thrombus on brain MRI in patients with hyper-acute ischemic stroke. 7th International Conference on Medical Imaging with Deep Learning (MIDL 2024), Medical Imaging with Deep Learning Foundation, Jul 2024, Paris, France. pp.657--671. ⟨hal-04939296⟩
0 Consultations
0 Téléchargements

Partager

More