DeSamba: Decoupled Spectral Adaptive Framework for 3D Multi-Sequence MRI Lesion Classification

2025-07-21

AI Summary

To address cross-modal feature coupling interference and insufficient spatial-frequency information utilization in 3D multi-sequence MRI lesion classification, this paper proposes a Decoupled Spectral-domain Adaptive Fusion (DSAF) framework. DSAF comprises a decoupled representation learning module and a spectral-domain adaptive modulation block, enabling self-reconstruction and cross-reconstruction decoupling of multi-sequence features while dynamically fusing spatial- and frequency-domain information according to lesion characteristics. Integrating the Mamba architecture, 3D CNNs, and a self-supervised reconstruction strategy, DSAF enhances feature discriminability and robustness. On the six-class spinal metastasis classification task, it achieves 62.10% Top-1 and 93.55% Top-3 accuracy. For spinal spondylitis classification, it attains internal and external validation AUCs of 74.75% and 73.88%, respectively—surpassing all existing state-of-the-art methods.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Magnetic Resonance Imaging (MRI) sequences provide rich spatial and frequency domain information, which is crucial for accurate lesion classification in medical imaging. However, effectively integrating multi-sequence MRI data for robust 3D lesion classification remains a challenge. In this paper, we propose DeSamba (Decoupled Spectral Adaptive Network and Mamba-Based Model), a novel framework designed to extract decoupled representations and adaptively fuse spatial and spectral features for lesion classification. DeSamba introduces a Decoupled Representation Learning Module (DRLM) that decouples features from different MRI sequences through self-reconstruction and cross-reconstruction, and a Spectral Adaptive Modulation Block (SAMB) within the proposed SAMNet, enabling dynamic fusion of spectral and spatial information based on lesion characteristics. We evaluate DeSamba on two clinically relevant 3D datasets. On a six-class spinal metastasis dataset (n=1,448), DeSamba achieves 62.10% Top-1 accuracy, 63.62% F1-score, 87.71% AUC, and 93.55% Top-3 accuracy on an external validation set (n=372), outperforming all state-of-the-art (SOTA) baselines. On a spondylitis dataset (n=251) involving a challenging binary classification task, DeSamba achieves 70.00%/64.52% accuracy and 74.75/73.88 AUC on internal and external validation sets, respectively. Ablation studies demonstrate that both DRLM and SAMB significantly contribute to overall performance, with over 10% relative improvement compared to the baseline. Our results highlight the potential of DeSamba as a generalizable and effective solution for 3D lesion classification in multi-sequence medical imaging.

Problem

• Integrating multi-sequence MRI data for 3D lesion classification
• Decoupling and fusing spatial and spectral MRI features
• Improving accuracy in medical imaging lesion diagnosis

Innovation

• Decoupled feature learning via self and cross-reconstruction
• Dynamic fusion of spectral and spatial features
• Mamba-based model for 3D lesion classification