A Recurrent Network for Segmenting the Thrombus on Brain MRI in Patients with Hyper-Acute Ischemic Stroke
Abstract
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 the Convolutional Long-Short-Term Memory (CLSTM) model. 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.
Cite
Text
Ibarra et al. "A Recurrent Network for Segmenting the Thrombus on Brain MRI in Patients with Hyper-Acute Ischemic Stroke." Proceedings of MIDL 2024, 2024.Markdown
[Ibarra et al. "A Recurrent Network for Segmenting the Thrombus on Brain MRI in Patients with Hyper-Acute Ischemic Stroke." Proceedings of MIDL 2024, 2024.](https://mlanthology.org/midl/2024/ibarra2024midl-recurrent/)BibTeX
@inproceedings{ibarra2024midl-recurrent,
title = {{A Recurrent Network for Segmenting the Thrombus on Brain MRI in Patients with Hyper-Acute Ischemic Stroke}},
author = {Ibarra, Sofia Vargas and Vigneron, Vincent Martin and Salicetti, Sonia Garcia and Maaref, Hichem and Kobold, Jonathan and Chausson, Nicolas and Lhermitte, Yann and Smadja, Didier},
booktitle = {Proceedings of MIDL 2024},
year = {2024},
pages = {657-671},
volume = {250},
url = {https://mlanthology.org/midl/2024/ibarra2024midl-recurrent/}
}