Enhancing 3T BOLD fMRI SNR using Unpaired 7T Data with Schr""odinger Bridge Diffusion

2025-04-01

AI Summary

Achieving high spatiotemporal resolution and signal-to-noise ratio (SNR) in 3T BOLD fMRI remains challenging, as it lags behind the performance of 7T systems—yet paired high-field/low-field fMRI data are rarely available in practice. Method: We propose an unsupervised domain translation framework based on the Schrödinger Bridge (SB) diffusion model, uniquely integrating unpaired cross-field fMRI alignment with Brain Disk geometric constraints—requiring neither shared subjects nor synchronized acquisitions. The method jointly models parametric inter-field registration and diffusion dynamics, validated functionally via population receptive field (pRF) fitting accuracy. Contribution/Results: Our approach significantly enhances SNR in 3T fMRI, yielding pRF model performance comparable to empirically acquired 7T data. Validation on both real retinotopic mapping datasets and synthetic benchmarks confirms robustness and fidelity across modalities.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

High spatial and temporal resolution, coupled with a strong signal-to-noise ratio (SNR), has made BOLD 7 Tesla fMRI an invaluable tool for understanding how the brain processes visual stimuli. However, the limited availability of 7T MRI systems means that most research relies on 3T MRI systems, which offer lower spatial and temporal resolution and SNR. This naturally raises the question: Can we enhance the spatiotemporal resolution and SNR of 3T BOLD fMRI data to approximate 7T quality? In this study, we propose a novel framework that aligns 7T and 3T fMRI data from different subjects and datasets in a shared parametric domain. We then apply an unpaired Brain Disk Schr""odinger Bridge diffusion model to enhance the spatiotemporal resolution and SNR of the 3T data. Our approach addresses the challenge of limited 7T data by improving the 3T scan quality. We demonstrate its effectiveness by testing it on two distinct fMRI retinotopy datasets (one 7T and one 3T), as well as synthetic data. The results show that our method significantly improves the SNR and goodness-of-fit of the population receptive field (pRF) model in the enhanced 3T data, making it comparable to 7T quality. The codes will be available at Github.

Problem

• Enhancing 3T fMRI SNR using unpaired 7T data
• Improving spatiotemporal resolution of 3T BOLD fMRI
• Bridging quality gap between 3T and 7T MRI systems

Innovation

• Unpaired 7T and 3T fMRI data alignment
• Schrödinger Bridge diffusion model application
• Enhanced 3T data comparable to 7T quality