DD-INR: Dynamics-Driven Implicit Neural Representation for Accelerated Whole-Brain Functional MRI Reconstruction

2026-06-09

AI Summary

This work addresses the challenge of reconstructing task-evoked blood-oxygen-level-dependent (BOLD) signals in highly accelerated functional MRI (fMRI), where conventional methods often sacrifice temporal fidelity for spatial accuracy, thereby failing to recover subtle neural activations. To overcome this limitation, the authors propose a novel reconstruction framework that decomposes fMRI data into static background and dynamic components, applying implicit neural representations (INRs) exclusively to the dynamic part. By integrating an incoherent time-varying undersampling strategy with tailored spatiotemporal priors, the method prioritizes the recovery of activation-related signals. This study presents the first application of dynamics-driven INRs to accelerated fMRI, significantly enhancing sensitivity and robustness to weak BOLD responses. Experiments on both simulated and in vivo data demonstrate superior performance over existing approaches in terms of image quality and accurate restoration of task activation patterns, enabling high-sensitivity fMRI with substantially reduced scan times.

Abstract

Accelerated acquisition of fMRI enables enhanced detection of neurovascular (BOLD) activity in the brain, but image reconstruction becomes challenging with high k-space undersampling: Task-evoked BOLD signals are small in magnitude, which traditional anatomical MRI reconstruction methods fail to recover, as they favor spatial accuracy over temporal fidelity. We present DD-INR, a Dynamics-Driven Implicit Neural Representation framework tailored for accelerated fMRI that benefits from incoherent time-varying sampling and a tailored spatiotemporal prior, outperforming traditional methods, demonstrated in simulation and in-vivo acquisition, both in terms of image quality and retrieval of activation patterns. DD-INR achieves this by splitting the fMRI data into a static background and a temporally varying dynamic component, representing only the dynamics with a dedicated INR, thereby focusing the model's capacity on activation-relevant changes while remaining compact. In general, DD-INR provides a promising framework for accelerated fMRI reconstruction, with the potential to improve the sensitivity and robustness of fMRI studies within practical scan time limits. The source code is available at https://github.com/JoosenLi/DD-INR.

Problem

• fMRI reconstruction
• k-space undersampling
• BOLD signal
• temporal fidelity
• accelerated MRI

Innovation

• Implicit Neural Representation
• accelerated fMRI
• spatiotemporal prior
• BOLD signal recovery
• dynamic component modeling