{"slug": "mosaik-motion-search-with-artificial-intelligence-in-k-space", "title": "MoSAIK: Motion Search with Artificial Intelligence in K-space", "summary": "A Transformer-based method called MoSAIK (Motion Search with AI in K-space) retrospectively estimates in-plane rigid-body motion directly from standard multi-coil T2-weighted turbo spin-echo brain MRI k-space, reporting rotation and translation estimation errors below 0.2 deg² and 0.2 pixels² mean squared error with Pearson correlations of approximately 0.96 and 0.94, according to the arXiv paper 2610.27156v1. Trained on motion-free clinical acquisitions with simulated motion and evaluated on four held-out subjects, MoSAIK produced information gain in more than 75% of slices in every in-vivo acquisition relative to uncorrected reconstructions, reducing visible artefacts while preserving anatomical detail. The method requires no motion-free reference, external navigators or additional acquisitions, feeding predicted parameters into model-based motion-aware reconstruction rather than synthesizing corrected images.", "body_md": "arXiv:2610.27156v1 Announce Type: new \nAbstract: Object:\n  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.\n  Materials and Methods:\n  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.\n  Results:\n  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.\n  Discussion:\n  Transformer-based k-space motion-trajectory estimation enables fast retrospective correction of standard clinical TSE brain MRI.", "url": "https://wpnews.pro/news/mosaik-motion-search-with-artificial-intelligence-in-k-space", "canonical_source": "https://www.machinebrief.com/news/mosaik-motion-search-with-artificial-intelligence-in-k-space-4wba", "published_at": "2026-10-21 04:00:00+00:00", "updated_at": "2026-10-21 04:50:06.416452+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "ai-research", "neural-networks"], "entities": ["MoSAIK", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/mosaik-motion-search-with-artificial-intelligence-in-k-space", "markdown": "https://wpnews.pro/news/mosaik-motion-search-with-artificial-intelligence-in-k-space.md", "text": "https://wpnews.pro/news/mosaik-motion-search-with-artificial-intelligence-in-k-space.txt", "jsonld": "https://wpnews.pro/news/mosaik-motion-search-with-artificial-intelligence-in-k-space.jsonld"}}