Deep learning for temporal super-resolution 4D Flow MRI

2025-01-15

AI Summary

Clinical 4D flow MRI is limited by insufficient temporal resolution, hindering accurate capture of rapid hemodynamic changes. This work proposes the first deep learning framework dedicated to pure temporal upsampling—enabling high-frame-rate flow reconstruction without prolonging scan time. We innovatively adapt a deep residual network to the 4D flow MRI temporal super-resolution task, departing from conventional spatial super-resolution paradigms and enabling physiologically plausible transient velocity sequence generation. Leveraging an enhanced 4DFlowNet architecture, the model is trained on patient-specific in-silico data and jointly validated on in-vivo acquisitions, with mean absolute error (MAE) and peak-flow phase correlation as primary metrics. On unseen simulated data, our method achieves an MAE of 1.0 cm/s—substantially outperforming linear (2.3 cm/s) and sinc interpolation (2.6 cm/s). In vivo results successfully reproduce peak-flow timing with excellent phase correlation, demonstrating clinical feasibility.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Planning, Routing, and Scheduling: Temporal Planning
• Computer Vision: Multi-modal Vision

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

4D Flow Magnetic Resonance Imaging (4D Flow MRI) is a non-invasive technique for volumetric, time-resolved blood flow quantification. However, apparent trade-offs between acquisition time, image noise, and resolution limit clinical applicability. In particular, in regions of highly transient flow, coarse temporal resolution can hinder accurate capture of physiologically relevant flow variations. To overcome these issues, post-processing techniques using deep learning have shown promising results to enhance resolution post-scan using so-called super-resolution networks. However, while super-resolution has been focusing on spatial upsampling, temporal super-resolution remains largely unexplored. The aim of this study was therefore to implement and evaluate a residual network for temporal super-resolution 4D Flow MRI. To achieve this, an existing spatial network (4DFlowNet) was re-designed for temporal upsampling, adapting input dimensions, and optimizing internal layer structures. Training and testing were performed using synthetic 4D Flow MRI data originating from patient-specific in-silico models, as well as using in-vivo datasets. Overall, excellent performance was achieved with input velocities effectively denoised and temporally upsampled, with a mean absolute error (MAE) of 1.0 cm/s in an unseen in-silico setting, outperforming deterministic alternatives (linear interpolation MAE = 2.3 cm/s, sinc interpolation MAE = 2.6 cm/s). Further, the network synthesized high-resolution temporal information from unseen low-resolution in-vivo data, with strong correlation observed at peak flow frames. As such, our results highlight the potential of utilizing data-driven neural networks for temporal super-resolution 4D Flow MRI, enabling high-frame-rate flow quantification without extending acquisition times beyond clinically acceptable limits.

Problem

• 4D Flow MRI
• Temporal Resolution
• Blood Flow Dynamics

Innovation

• Deep Learning
• Residual Network
• 4D Flow MRI Resolution Enhancement