ReBrain: Brain MRI Reconstruction from Sparse CT Slice via Retrieval-Augmented Diffusion

2025-11-21

AI Summary

Synthesizing high-fidelity brain MRI from low-dose CT is challenging due to sparse slice sampling and low spatial resolution. Method: We propose a retrieval-augmented Brownian bridge diffusion framework comprising: (1) a structural-aware CT prior database with a fine-tuned retrieval model to identify anatomically similar slices; (2) ControlNet-guided 2D MRI synthesis, augmented with spherical linear interpolation (Slerp) to ensure robustness upon retrieval failure; and (3) joint slice-level generation and 3D consistency regularization for anatomically coherent cross-modal reconstruction. Results: Evaluated on SynthRAD2023 and BraTS, our method achieves state-of-the-art performance in PSNR, SSIM, and anatomical structure preservation—outperforming existing approaches. It provides a reliable MRI surrogate for brain disorder diagnosis in MRI-absent clinical settings.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Magnetic Resonance Imaging (MRI) plays a crucial role in brain disease diagnosis, but it is not always feasible for certain patients due to physical or clinical constraints. Recent studies attempt to synthesize MRI from Computed Tomography (CT) scans; however, low-dose protocols often result in highly sparse CT volumes with poor through-plane resolution, making accurate reconstruction of the full brain MRI volume particularly challenging. To address this, we propose ReBrain, a retrieval-augmented diffusion framework for brain MRI reconstruction. Given any 3D CT scan with limited slices, we first employ a Brownian Bridge Diffusion Model (BBDM) to synthesize MRI slices along the 2D dimension. Simultaneously, we retrieve structurally and pathologically similar CT slices from a comprehensive prior database via a fine-tuned retrieval model. These retrieved slices are used as references, incorporated through a ControlNet branch to guide the generation of intermediate MRI slices and ensure structural continuity. We further account for rare retrieval failures when the database lacks suitable references and apply spherical linear interpolation to provide supplementary guidance. Extensive experiments on SynthRAD2023 and BraTS demonstrate that ReBrain achieves state-of-the-art performance in cross-modal reconstruction under sparse conditions.

Problem

• Reconstructing full brain MRI volumes from sparse CT slices
• Addressing poor through-plane resolution in low-dose CT scans
• Overcoming retrieval failures when prior database lacks references

Innovation

• Retrieval-augmented diffusion framework for MRI reconstruction
• Brownian Bridge Diffusion Model synthesizes MRI from CT
• ControlNet integrates retrieved references for structural continuity