Pattern-Aware Diffusion Synthesis of fMRI/dMRI with Tissue and Microstructural Refinement

2025-11-07

AI Summary

To address the modality gap in cross-modal synthesis between fMRI and dMRI—arising from fundamental physical differences between BOLD and diffusion signals—and the limitation of existing methods in neglecting disease-relevant neuroanatomical patterns, this paper proposes a pattern-aware bimodal 3D diffusion model. The model integrates a tissue-refined network with a microstructure-guided refinement module, enabling explicit embedding of pathology-informed anatomical priors through bimodal collaborative learning, thereby enhancing structural fidelity and pathological sensitivity of synthesized images. Trained end-to-end on multicenter OASIS-3 and ADNI datasets, the model achieves PSNRs of 29.83 dB for fMRI and 30.00 dB for dMRI synthesis, with a maximum clinical diagnostic accuracy of 67.92%, significantly outperforming state-of-the-art methods.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Magnetic resonance imaging (MRI), especially functional MRI (fMRI) and diffusion MRI (dMRI), is essential for studying neurodegenerative diseases. However, missing modalities pose a major barrier to their clinical use. Although GAN- and diffusion model-based approaches have shown some promise in modality completion, they remain limited in fMRI-dMRI synthesis due to (1) significant BOLD vs. diffusion-weighted signal differences between fMRI and dMRI in time/gradient axis, and (2) inadequate integration of disease-related neuroanatomical patterns during generation. To address these challenges, we propose PDS, introducing two key innovations: (1) a pattern-aware dual-modal 3D diffusion framework for cross-modality learning, and (2) a tissue refinement network integrated with a efficient microstructure refinement to maintain structural fidelity and fine details. Evaluated on OASIS-3, ADNI, and in-house datasets, our method achieves state-of-the-art results, with PSNR/SSIM scores of 29.83 dB/90.84% for fMRI synthesis (+1.54 dB/+4.12% over baselines) and 30.00 dB/77.55% for dMRI synthesis (+1.02 dB/+2.2%). In clinical validation, the synthesized data show strong diagnostic performance, achieving 67.92%/66.02%/64.15% accuracy (NC vs. MCI vs. AD) in hybrid real-synthetic experiments. Code is available in href{https://github.com/SXR3015/PDS}{PDS GitHub Repository}

Problem

• Synthesizing missing fMRI and dMRI modalities for clinical use
• Addressing signal differences between fMRI and dMRI in synthesis
• Integrating disease-related neuroanatomical patterns during generation

Innovation

• Pattern-aware dual-modal 3D diffusion framework
• Tissue refinement network for structural fidelity
• Microstructure refinement for fine detail preservation