Distributional Deep Learning for Super-Resolution of 4D Flow MRI under Domain Shift

TL;DR

This paper introduces a distributional deep learning framework to improve super-resolution of 4D Flow MRI, especially under domain shift conditions, by enhancing robustness and generalization from simulated to real clinical data.

Contribution

The paper presents a novel distributional deep learning approach for super-resolution of 4D Flow MRI that addresses domain shift, with theoretical analysis and superior real-world performance.

Findings

Outperforms traditional methods in real data scenarios

Improves robustness to domain shift in MRI super-resolution

Theoretically grounded distributional estimators

Abstract

Super-resolution is widely used in medical imaging to enhance low-quality data, reducing scan time and improving abnormality detection. Conventional super-resolution approaches typically rely on paired datasets of downsampled and original high resolution images, training models to reconstruct high resolution images from their artificially degraded counterparts. However, in real-world clinical settings, low resolution data often arise from acquisition mechanisms that differ significantly from simple downsampling. As a result, these inputs may lie outside the domain of the training data, leading to poor model generalization due to domain shift. To address this limitation, we propose a distributional deep learning framework that improves model robustness and domain generalization. We develop this approch for enhancing the resolution of 4D Flow MRI (4DF). This is a novel imaging modality that captures hemodynamic flow velocity and clinically relevant metrics such as vessel wall stress. These metrics are critical for assessing aneurysm rupture risk. Our model is initially trained on high resolution computational fluid dynamics (CFD) simulations and their downsampled counterparts. It is then fine-tuned on a small, harmonized dataset of paired 4D Flow MRI and CFD samples. We derive the theoretical properties of our distributional estimators and demonstrate that our framework significantly outperforms traditional deep learning approaches through real data applications. This highlights the effectiveness of distributional learning in addressing domain shift and improving super-resolution performance in clinically realistic scenarios.