Saved in:
Bibliographic Details
Main Authors: Fedorov, Alex, Bu, Yutong, Hu, Xiao, Rorden, Chris, Plis, Sergey
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.05531
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929747424968704
author Fedorov, Alex
Bu, Yutong
Hu, Xiao
Rorden, Chris
Plis, Sergey
author_facet Fedorov, Alex
Bu, Yutong
Hu, Xiao
Rorden, Chris
Plis, Sergey
contents Efficient and accurate whole-brain lesion segmentation remains a challenge in medical image analysis. In this work, we revisit MeshNet, a parameter-efficient segmentation model, and introduce a novel multi-scale dilation pattern with an encoder-decoder structure. This innovation enables capturing broad contextual information and fine-grained details without traditional downsampling, upsampling, or skip-connections. Unlike previous approaches processing subvolumes or slices, we operate directly on whole-brain $256^3$ MRI volumes. Evaluations on the Aphasia Recovery Cohort (ARC) dataset demonstrate that MeshNet achieves superior or comparable DICE scores to state-of-the-art architectures such as MedNeXt and U-MAMBA at 1/1000th of parameters. Our results validate MeshNet's strong balance of efficiency and performance, making it particularly suitable for resource-limited environments such as web-based applications and opening new possibilities for the widespread deployment of advanced medical image analysis tools.
format Preprint
id arxiv_https___arxiv_org_abs_2503_05531
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle State-of-the-Art Stroke Lesion Segmentation at 1/1000th of Parameters
Fedorov, Alex
Bu, Yutong
Hu, Xiao
Rorden, Chris
Plis, Sergey
Image and Video Processing
Computer Vision and Pattern Recognition
Efficient and accurate whole-brain lesion segmentation remains a challenge in medical image analysis. In this work, we revisit MeshNet, a parameter-efficient segmentation model, and introduce a novel multi-scale dilation pattern with an encoder-decoder structure. This innovation enables capturing broad contextual information and fine-grained details without traditional downsampling, upsampling, or skip-connections. Unlike previous approaches processing subvolumes or slices, we operate directly on whole-brain $256^3$ MRI volumes. Evaluations on the Aphasia Recovery Cohort (ARC) dataset demonstrate that MeshNet achieves superior or comparable DICE scores to state-of-the-art architectures such as MedNeXt and U-MAMBA at 1/1000th of parameters. Our results validate MeshNet's strong balance of efficiency and performance, making it particularly suitable for resource-limited environments such as web-based applications and opening new possibilities for the widespread deployment of advanced medical image analysis tools.
title State-of-the-Art Stroke Lesion Segmentation at 1/1000th of Parameters
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.05531