MoSAIK: Motion Search with Artificial Intelligence in K-space
Abstract
Object: Motion artefacts remain a major limitation of routine two-dimensional T2-weighted turbo spin-echo (TSE) brain MRI, reducing diagnostic confidence and often necessitating repeated acquisitions. MoSAIK (Motion Search with AI in K-space) retrospectively estimates in-plane rigid-body motion directly from standard multi-coil TSE k-space. Without a motion-free reference, external navigators, or additional acquisitions, MoSAIK infers motion from k-space self-consistency violations and incorporates predicted parameters into model-based, motion-aware reconstruction rather than synthesizing corrected images. Materials and Methods: Selected complex coil data were represented as log-magnitude and phase channels. A decoupled convolutional embedding and Transformer encoder estimated phase-encoding-line-indexed rotation and two-dimensional translations. Training used motion-free clinical acquisitions with simulated motion; four held-out subjects were evaluated. Component-wise errors assessed estimation; SSIM and PSNR in silico and entropy focus criterion (EFC) in vivo assessed reconstruction. Results: Motion estimation errors were low for both rotation (MSE < 0.2 deg2) and translation (MSE < 0.2 pixels2), with Pearson correlations of approximately 0.96 and 0.94, respectively. In every in-vivo acquisition, more than 75% of slices showed information gain relative to uncorrected reconstructions. MoSAIK reduced visible artefacts while preserving anatomical detail. Discussion: Transformer-based k-space motion-trajectory estimation enables fast retrospective correction of standard clinical TSE brain MRI.
Related
- Relaxation of a Vlasov gas to an inhomogeneous state due to phase space mixing in an axisymmetric potential: A Newtonian analogy of the Kerr orbital motion
- Continuous Space-Time Finite Element Simulations for Biomedical Flows with Large Boundary Motion and Time-Periodicity
- Wide parameter-space O3 search for continuous gravitational waves from unknown neutron stars in binary systems
- A Sign in Space part 2, a new simulated First Contact scenario engaging the global scientific SETI network, and integrating deep space technologies into SETI search
- \texttt{cWB-space}: A time-frequency transient-search pipeline for LISA
BibTeX
@misc{dabrowski2026mosaik,
title = {{MoSAIK: Motion Search with Artificial Intelligence in K-space}},
author = {Oscar Dabrowski and Abijith Jagannath and Frédéric Grouiller and Sébastien Courvoisier},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.27156},
eprint = {2610.27156},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.27156},
}
RIS
TY - UNPB
TI - MoSAIK: Motion Search with Artificial Intelligence in K-space
AU - Oscar Dabrowski
AU - Abijith Jagannath
AU - Frédéric Grouiller
AU - Sébastien Courvoisier
PY - 2026
DA - 2026/10/19
PB - arXiv
DO - 10.48550/arxiv.2610.27156
UR - https://arxiv.org/abs/2610.27156
AB - Object: Motion artefacts remain a major limitation of routine two-dimensional T2-weighted turbo spin-echo (TSE) brain MRI, reducing diagnostic confidence and often necessitating repeated acquisitions. MoSAIK (Motion Search with AI in K-space) retrospectively estimates in-plane rigid-body motion directly from standard multi-coil TSE k-space. Without a motion-free reference, external navigators, or additional acquisitions, MoSAIK infers motion from k-space self-consistency violations and incorporates predicted parameters into model-based, motion-aware reconstruction rather than synthesizing corrected images. Materials and Methods: Selected complex coil data were represented as log-magnitude and phase channels. A decoupled convolutional embedding and Transformer encoder estimated phase-encoding-line-indexed rotation and two-dimensional translations. Training used motion-free clinical acquisitions with simulated motion; four held-out subjects were evaluated. Component-wise errors assessed estimation; SSIM and PSNR in silico and entropy focus criterion (EFC) in vivo assessed reconstruction. Results: Motion estimation errors were low for both rotation (MSE < 0.2 deg^2) and translation (MSE < 0.2 pixels^2), with Pearson correlations of approximately 0.96 and 0.94, respectively. In every in-vivo acquisition, more than 75% of slices showed information gain relative to uncorrected reconstructions. MoSAIK reduced visible artefacts while preserving anatomical detail. Discussion: Transformer-based k-space motion-trajectory estimation enables fast retrospective correction of standard clinical TSE brain MRI.
ER -