Versatile and Efficient Medical Image Super-Resolution Via Frequency-Gated Mamba

2025-10-31

AI Summary

Medical image super-resolution must balance diagnostic accuracy with computational efficiency, yet existing methods struggle to jointly model long-range anatomical structures and fine-grained frequency-domain details under low computational overhead. To address this, we propose a frequency-aware lightweight Mamba architecture. Our method introduces the Global Attention-State Mixer (GASM) module, which synergistically integrates state-space modeling with dual spatial-channel attention, and the FFT-guided Pyramid Frequency Fusion Module (PFFM), enabling efficient global dependency capture and multi-scale high-frequency detail reconstruction. The resulting model contains only 0.75 million parameters. Extensive experiments across five modalities—ultrasound, optical coherence tomography (OCT), MRI, CT, and endoscopy—demonstrate significant PSNR/SSIM improvements over state-of-the-art CNN- and Transformer-based approaches, achieving both high reconstruction fidelity and strong cross-modality generalization.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• 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

Medical image super-resolution (SR) is essential for enhancing diagnostic accuracy while reducing acquisition cost and scanning time. However, modeling both long-range anatomical structures and fine-grained frequency details with low computational overhead remains challenging. We propose FGMamba, a novel frequency-aware gated state-space model that unifies global dependency modeling and fine-detail enhancement into a lightweight architecture. Our method introduces two key innovations: a Gated Attention-enhanced State-Space Module (GASM) that integrates efficient state-space modeling with dual-branch spatial and channel attention, and a Pyramid Frequency Fusion Module (PFFM) that captures high-frequency details across multiple resolutions via FFT-guided fusion. Extensive evaluations across five medical imaging modalities (Ultrasound, OCT, MRI, CT, and Endoscopic) demonstrate that FGMamba achieves superior PSNR/SSIM while maintaining a compact parameter footprint ($<$0.75M), outperforming CNN-based and Transformer-based SOTAs. Our results validate the effectiveness of frequency-aware state-space modeling for scalable and accurate medical image enhancement.

Problem

• Enhancing medical image resolution for improved diagnostic accuracy
• Modeling long-range anatomical structures with low computational overhead
• Capturing fine-grained frequency details across multiple imaging modalities

Innovation

• Frequency-aware gated state-space model for medical SR
• Gated attention-enhanced state-space module with dual-branch attention
• Pyramid frequency fusion module using FFT-guided multi-resolution fusion