MS-UMamba: An Improved Vision Mamba Unet for Fetal Abdominal Medical Image Segmentation

TL;DR

MS-UMamba is a hybrid convolutional-mamba model that improves fetal ultrasound image segmentation by balancing local and global features, utilizing multi-scale fusion and attention mechanisms, and demonstrating superior performance on a specialized dataset.

Contribution

The paper introduces MS-UMamba, combining CNN and Mamba modules with multi-scale fusion and attention, advancing fetal ultrasound segmentation accuracy.

Findings

Enhanced segmentation accuracy on fetal ultrasound images.

Effective integration of local and global features.

Superior performance compared to existing methods.

Abstract

Recently, Mamba-based methods have become popular in medical image segmentation due to their lightweight design and long-range dependency modeling capabilities. However, current segmentation methods frequently encounter challenges in fetal ultrasound images, such as enclosed anatomical structures, blurred boundaries, and small anatomical structures. To address the need for balancing local feature extraction and global context modeling, we propose MS-UMamba, a novel hybrid convolutional-mamba model for fetal ultrasound image segmentation. Specifically, we design a visual state space block integrated with a CNN branch (SS-MCAT-SSM), which leverages Mamba's global modeling strengths and convolutional layers' local representation advantages to enhance feature learning. In addition, we also propose an efficient multi-scale feature fusion module that integrates spatial attention mechanisms, which Integrating feature information from different layers enhances the feature representation ability of the model. Finally, we conduct extensive experiments on a non-public dataset, experimental results demonstrate that MS-UMamba model has excellent performance in segmentation performance.