Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Resolution- and Stimulus-agnostic Super-Resolution of Ultra-High-Field Functional MRI: Application to Visual Studies

Hongwei Bran Li, Matthew S. Rosen, Shahin Nasr, Juan Eugenio Iglesias

TL;DR

The study addresses the challenge of achieving high spatial fidelity in ultra-high-field fMRI without extending scan times by introducing a 3D deep learning SR framework with resolution-agnostic augmentation. The method learns a LR-to-HR interpolation that generalizes across voxel sizes and tasks, using a compound L1+SSIM loss within a 3D FCN and applies frame-by-frame to produce 4D SR fMRI. It demonstrates recovery of fine-scale, interdigitated motion-selective sites in early visual areas from 2–3 mm data, with robust cross-subject and cross-task performance and improved gray–white matter delineation. This approach has practical implications for faster acquisition and more precise mesoscale functional mapping in both research and clinical settings.

Abstract

High-resolution fMRI provides a window into the brain's mesoscale organization. Yet, higher spatial resolution increases scan times, to compensate for the low signal and contrast-to-noise ratio. This work introduces a deep learning-based 3D super-resolution (SR) method for fMRI. By incorporating a resolution-agnostic image augmentation framework, our method adapts to varying voxel sizes without retraining. We apply this innovative technique to localize fine-scale motion-selective sites in the early visual areas. Detection of these sites typically requires a resolution higher than 1 mm isotropic, whereas here, we visualize them based on lower resolution (2-3mm isotropic) fMRI data. Remarkably, the super-resolved fMRI is able to recover high-frequency detail of the interdigitated organization of these sites (relative to the color-selective sites), even with training data sourced from different subjects and experimental paradigms -- including non-visual resting-state fMRI, underscoring its robustness and versatility. Quantitative and qualitative results indicate that our method has the potential to enhance the spatial resolution of fMRI, leading to a drastic reduction in acquisition time.

Resolution- and Stimulus-agnostic Super-Resolution of Ultra-High-Field Functional MRI: Application to Visual Studies

TL;DR

The study addresses the challenge of achieving high spatial fidelity in ultra-high-field fMRI without extending scan times by introducing a 3D deep learning SR framework with resolution-agnostic augmentation. The method learns a LR-to-HR interpolation that generalizes across voxel sizes and tasks, using a compound L1+SSIM loss within a 3D FCN and applies frame-by-frame to produce 4D SR fMRI. It demonstrates recovery of fine-scale, interdigitated motion-selective sites in early visual areas from 2–3 mm data, with robust cross-subject and cross-task performance and improved gray–white matter delineation. This approach has practical implications for faster acquisition and more precise mesoscale functional mapping in both research and clinical settings.

Abstract

High-resolution fMRI provides a window into the brain's mesoscale organization. Yet, higher spatial resolution increases scan times, to compensate for the low signal and contrast-to-noise ratio. This work introduces a deep learning-based 3D super-resolution (SR) method for fMRI. By incorporating a resolution-agnostic image augmentation framework, our method adapts to varying voxel sizes without retraining. We apply this innovative technique to localize fine-scale motion-selective sites in the early visual areas. Detection of these sites typically requires a resolution higher than 1 mm isotropic, whereas here, we visualize them based on lower resolution (2-3mm isotropic) fMRI data. Remarkably, the super-resolved fMRI is able to recover high-frequency detail of the interdigitated organization of these sites (relative to the color-selective sites), even with training data sourced from different subjects and experimental paradigms -- including non-visual resting-state fMRI, underscoring its robustness and versatility. Quantitative and qualitative results indicate that our method has the potential to enhance the spatial resolution of fMRI, leading to a drastic reduction in acquisition time.
Paper Structure (9 sections, 1 equation, 4 figures)

This paper contains 9 sections, 1 equation, 4 figures.

Table of Contents

  1. INTRODUCTION
  2. Methods
  3. Experiments and Results
  4. Image Acquisition
  5. Experimental setup
  6. Qualitative Results of 3D Frames
  7. Quantitative Results of fMRI Analysis
  8. Summary and Conclusion
  9. Acknowledgement

Figures (4)

  • Figure 1: Training overview. We parse the 4D (3D+t) fMRI data to single-time-point 3D frames. We use a function $\phi_{1}$ to perform random affine and non-linear transformations, and random contrast adjustment. We then downgrade the augmented HR frames by another function $\phi_{2}$ to perform downsampling to random lower resolutions (including a low-pass antialias filter), adding noise, and linear interpolation to the original size.
  • Figure 2: Image quality enhancement by applying the SR method to low-resolution images. Gray-white matter segregation is more apparent in super-resolved images compared to downsampled images.
  • Figure 3: The application of the SR method improves the localization of motion-selective activity maps from low-resolution fMRI. A) Localization of motion-selective sites across V2, V3, V3A, and V4 based on the original high-resolution fMRI. B-C) Localization of the same sites based on downsampled data. Fine-scale sites are either absent or fused, causing overestimation in the size of the selective sites. D-E) Localization of motion-selective sites based on super-resolved images. The fine-scale motion-selective sites are mostly recovered in these maps. In all panels, dashed black lines indicate the borders of visual areas, defined retinotopically.
  • Figure 4: Consistency between original motion-selectivity maps and those from downsampled, super-resolved images are shown in two panels for 2mm and 3mm resolution datasets, respectively. The "consistency index" represents the correlation between regenerated and original motion-selectivity maps, subtracted by the correlation with an independent high-resolution color-selectivity map. (VS=visual stimuli, RS=resting state.)