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

2025-06-14

AI Summary

To address challenges in fetal ultrasound image segmentation—including enclosed anatomical structures, ambiguous boundaries, and difficulty segmenting small targets—this paper proposes a lightweight U-shaped network. The core innovation is the SS-MCAT-SSM module, the first visual state-space module integrating convolutional and Mamba architectures to jointly model local details and global contextual dependencies. Additionally, a spatial-attention-guided multi-scale feature fusion mechanism is introduced to enhance discriminative representation of critical regions. Evaluated on a private fetal abdominal ultrasound dataset, the method achieves significant improvements in segmenting small structures (e.g., gallbladder, renal pelvis) and objects with ill-defined boundaries, yielding an average Dice coefficient gain of 4.2%. Compared to mainstream Transformer-based baselines, the model reduces parameter count by 68%, achieving both superior accuracy and deployment efficiency.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

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.

Problem

• Improves fetal ultrasound image segmentation accuracy
• Balances local feature extraction and global context modeling
• Addresses blurred boundaries and small anatomical structures

Innovation

• Hybrid convolutional-mamba model for segmentation
• SS-MCAT-SSM block combines Mamba and CNN
• Multi-scale feature fusion with spatial attention