MMRINet: Efficient Mamba-Based Segmentation with Dual-Path Refinement for Low-Resource MRI Analysis

TL;DR

MMRINet is a lightweight, efficient MRI segmentation model that uses novel linear-complexity attention and feature refinement modules, achieving high accuracy in resource-limited settings.

Contribution

This work introduces MMRINet, a novel low-resource MRI segmentation architecture with dual-path refinement and linear attention, outperforming existing models in efficiency and accuracy.

Findings

Achieves 0.752 Dice score on BraTS-Lighthouse SSA 2025

Uses only ~2.5 million parameters, suitable for low-resource environments

Demonstrates efficient volumetric segmentation with high accuracy

Abstract

Automated brain tumor segmentation in multi-parametric MRI remains challenging in resource-constrained settings where deep 3D networks are computationally prohibitive. We propose MMRINet, a lightweight architecture that replaces quadratic-complexity attention with linear-complexity Mamba state-space models for efficient volumetric context modeling. Novel Dual-Path Feature Refinement (DPFR) modules maximize feature diversity without additional data requirements, while Progressive Feature Aggregation (PFA) enables effective multi-scale fusion. In the BraTS-Lighthouse SSA 2025, our model achieves strong performance with an average Dice score of (0.752) and an average HD95 of (12.23) with only ~2.5M parameters, demonstrating efficient and accurate segmentation suitable for low-resource clinical environments. Our GitHub repository can be accessed here: github.com/BioMedIA-MBZUAI/MMRINet.