State-of-the-Art Stroke Lesion Segmentation at 1/1000th of Parameters

2025-03-07

AI Summary

Balancing efficiency and accuracy in whole-brain stroke lesion segmentation remains challenging under resource-constrained settings. Method: We propose a lightweight 3D dilated convolutional encoder-decoder architecture that processes full-resolution 256³ brain MRI volumes end-to-end—eliminating conventional downsampling/upsampling, skip connections, and sub-voxel patching. A novel multi-scale dilated convolution pattern enables simultaneous modeling of global context and fine-grained local structures within a single forward pass. Contribution/Results: Our model achieves only ~0.1% of the parameters of state-of-the-art methods (e.g., MedNeXt, U-MAMBA) while attaining comparable Dice scores on the ARC dataset. It significantly accelerates inference and enhances deployment flexibility, having been successfully integrated into a web-based lightweight application. This provides scalable, AI-powered diagnostic support for primary healthcare settings.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

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.

Problem

• Efficient whole-brain stroke lesion segmentation challenge
• Novel multi-scale dilation pattern for MRI volumes
• MeshNet achieves high accuracy with minimal parameters

Innovation

• MeshNet with multi-scale dilation pattern
• Encoder-decoder structure without downsampling
• Whole-brain MRI volume processing