Revisiting Global Token Mixing in Task-Dependent MRI Restoration: Insights from Minimal Gated CNN Baselines

TL;DR

This study evaluates the effectiveness of global token mixing in MRI restoration tasks, revealing that its benefits are highly dependent on the specific task and underlying physics, with local models often performing competitively.

Contribution

The paper provides a controlled comparison between local gated CNNs and global token-mixing models across different MRI restoration tasks, highlighting when global mixing is advantageous.

Findings

Global token mixing offers limited benefits in accelerated MRI reconstruction with strong data consistency.

Local gated models perform competitively in MRI super-resolution where low-frequency data are preserved.

Token-mixing models excel in denoising tasks with spatially heteroscedastic noise.

Abstract

Global token mixing, implemented via self-attention or state-space sequence models, has become a popular model design choice for MRI restoration. However, MRI restoration tasks differ substantially in how their degradations vary over image and k-space domains, and in the degree to which global coupling is already imposed by physics-driven data consistency terms. In this work, we ask the question whether global token mixing is actually beneficial in each individual task across three representative settings: accelerated MRI reconstruction with explicit data consistency, MRI super-resolution with k-space center cropping, and denoising of clinical carotid MRI data with spatially heteroscedastic noise. To reduce confounding factors, we establish a controlled testbed comparing a minimal local gated CNN and its large-field variant, benchmarking them directly against state-of-the-art global models under aligned training and evaluation protocols. For accelerated MRI reconstruction, the minimal unrolled gated-CNN baseline is already highly competitive compared to recent token-mixing approaches in public reconstruction benchmarks, suggesting limited additional benefits when the forward model and data-consistency steps provide strong global constraints. For super-resolution, where low-frequency k-space data are largely preserved by the controlled low-pass degradation, local gated models remain competitive, and a lightweight large-field variant yields only modest improvements. In contrast, for denoising with pronounced spatially heteroscedastic noise, token-mixing models achieve the strongest overall performance, consistent with the need to estimate spatially varying reliability. In conclusion, our results demonstrate that the utility of global token mixing in MRI restoration is task-dependent, and it should be tailored to the underlying imaging physics and degradation structure.