High‐resolution multi‐shot diffusion‐weighted MRI combining markerless prospective motion correction and locally low‐rank constrained reconstruction

Abstract

Subject head motion is a major challenge in DWI, leading to image blurring, signal losses, and biases in the estimated diffusion parameters. Here, we investigate a combined application of prospective motion correction and spatial-angular locally low-rank constrained reconstruction to obtain robust, multi-shot, high-resolution diffusion-weighted MRI under substantial motion.Methods: Single-shot EPI with retrospective motion correction can mitigate motion artifacts and resolve any mismatching of gradient encoding orientations; however, it is limited by low spatial resolution and image distortions. Multi-shot acquisition strategies could achieve higher resolution and image fidelity but increase the vulnerability to motion artifacts and phase variations related to cardiac pulsations from shot to shot. We use prospective motion correction with optical markerless motion tracking to remove artifacts and reduce image blurring due to bulk motion, combined with locally low-rank regularization to correct for remaining artifacts due to shot-to-shot phase variations. Results:The approach was evaluated on healthy adult volunteers at 3 Tesla under different motion patterns. In multi-shot DWI, image blurring due to motion with 20 mm translations and 30 • rotations was successfully removed by prospective motion correction, and aliasing artifacts caused by shot-to-shot phase variations were addressed by locally low-rank regularization. The ability of prospective motion correction to preserve the orientational information in DTI without requiring a reorientation of the b-matrix is highlighted. Conclusion:The described technique is proved to hold valuable potential for mapping brain diffusivity and connectivity at high resolution for studies in subjects/cohorts where motion is common, including neonates, pediatrics, and patients with neurological disorders.

Motion compensated structured low‐rank reconstruction for 3D multi‐shot EPI

“…One class of methods relies on additional MRI navigators or external hardware (e.g., optical tracking systems) to obtain motion estimates, which can be used for either prospective or retrospective motion correction. [44][45][46][47][48][49] The other type of motion correction methods uses data-driven approaches to estimate motion parameters directly from the imaging data. [50][51][52][53][54][55][56] The methods AMUSE 52 and SENDIMENT 56 estimate inter-shot macroscale motion and motion induced phase errors for 2D multi-shot dMRI from SENSE reconstructed shot images.…”

“…Previous intra‐volume motion correction methods can be broadly divided into two categories depending on how motion parameters are estimated. One class of methods relies on additional MRI navigators or external hardware (e.g., optical tracking systems) to obtain motion estimates, which can be used for either prospective or retrospective motion correction 44–49 . The other type of motion correction methods uses data‐driven approaches to estimate motion parameters directly from the imaging data 50–56 .…”

Motion‐compensated diffusion encoding in multi‐shot human brain acquisitions: Insights using high‐performance gradients

“…A further disadvantage of these methods is computation time, which limits their use. 32,33 An alternative means of combatting phase variability in multi-shot diffusion experiments is to eliminate the cause of shot-to-shot phase differences, that is, to null the moments of the diffusion encoding gradients. Diffusion gradient shapes with nulled higher-order moments incur no phase accrual for equivalent-order time derivatives of motion, thereby reducing shot-to-shot phase differences, but encode diffusion less efficiently and require longer TEs (for a given b-value), thereby reducing SNR.…”

“…A drawback of these increasingly complex reconstruction strategies is that the assumed or enforced smoothness of motion‐induced phase inherently restricts attainable image accuracy, especially at higher interleaf factors for which undersampling impedes the estimation or inference of phase in each shot. A further disadvantage of these methods is computation time, which limits their use 32,33 …”

Markerless head motion tracking and event-by-event correction in brain PET

“…The core principle of MLMT involves employing computer vision techniques for robust feature matching, enabling the computation of changes in head position accurately. In the past, MLMT has been successfully applied in MRI motion tracking (Kyme et al 2020 , Chen et al 2023 ). Over time, several markerless motion tracking methods for PET have been proposed.…”