Enhanced Portable Ultra Low-Field Diffusion Tensor Imaging with Bayesian Artifact Correction and Deep Learning-Based Super-Resolution

2026-02-11

AI Summary

This study addresses the challenges of portable ultra-low-field diffusion tensor imaging (ULF DTI), which suffers from low spatial and angular resolution, poor signal-to-noise ratio, and cross-domain artifacts that hinder accurate white matter fiber reconstruction. To overcome these limitations, the authors propose a nine-direction single-shell ULF DTI acquisition protocol integrated with an angle-aware Bayesian bias field correction algorithm and a generalizable deep learning-based super-resolution model, DiffSR. Notably, DiffSR requires no retraining to generalize across datasets, effectively aligning ULF DTI with high-field DTI and substantially improving the fidelity of recovered white matter microstructural information. Experiments on both synthetically downsampled and real paired scans demonstrate that DiffSR significantly enhances the agreement of DTI metrics with those from non-degraded data, thereby supporting clinical applications such as Alzheimer’s disease classification.

Technology Category

• Machine Learning: Learning with Manifolds
• Computer Vision: Diffusion Models for Vision
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Portable, ultra-low-field (ULF) magnetic resonance imaging has the potential to expand access to neuroimaging but currently suffers from coarse spatial and angular resolutions and low signal-to-noise ratios. Diffusion tensor imaging (DTI), a sequence tailored to detect and reconstruct white matter tracts within the brain, is particularly prone to such imaging degradation due to inherent sequence design coupled with prolonged scan times. In addition, ULF DTI scans exhibit artifacting that spans both the space and angular domains, requiring a custom modelling algorithm for subsequent correction. We introduce a nine-direction, single-shell ULF DTI sequence, as well as a companion Bayesian bias field correction algorithm that possesses angular dependence and convolutional neural network-based superresolution algorithm that is generalizable across DTI datasets and does not require re-training (''DiffSR''). We show through a synthetic downsampling experiment and white matter assessment in real, matched ULF and high-field DTI scans that these algorithms can recover microstructural and volumetric white matter information at ULF. We also show that DiffSR can be directly applied to white matter-based Alzheimers disease classification in synthetically degraded scans, with notable improvements in agreement between DTI metrics, as compared to un-degraded scans. We freely disseminate the Bayesian bias correction algorithm and DiffSR with the goal of furthering progress on both ULF reconstruction methods and general DTI sequence harmonization. We release all code related to DiffSR for $\href{https://github.com/markolchanyi/DiffSR}{public \space use}$.

Problem

• ultra-low-field MRI
• diffusion tensor imaging
• image artifacts
• low resolution
• white matter

Innovation

• ultra-low-field MRI
• diffusion tensor imaging
• Bayesian artifact correction
• deep learning super-resolution
• DiffSR