Home  | Publications | ZSP+24b

Reconstruction-Free Segmentation From Undersampled K-Space Using Transformers

MCML Authors

Abstract

Motivation: High acceleration factors place a limit on MRI image reconstruction. This limit is extended to segmentation models when treating these as subsequent independent processes.<br>Goal(s): Our goal is to produce segmentations directly from sparse k-space measurements without the need for intermediate image reconstruction.<br>Approach: We employ a transformer architecture to encode global k-space information into latent features. The produced latent vectors condition queried coordinates during decoding to generate segmentation class probabilities.<br>Results: The model is able to produce better segmentations across high acceleration factors than image-based segmentation baselines.<br>Impact: Cardiac segmentation directly from undersampled k-space samples circumvents the need for an intermediate image reconstruction step. This allows the potential to assess myocardial structure and function on higher acceleration factors than methods that rely on images as input.

inproceedings

ISMRM 2024

International Society for Magnetic Resonance in Medicine Annual Meeting. Singapore, May 04-09, 2024.

Authors

Y. Zhang • N. Stolt-Ansó • J. Pan • W. Huang • K. Hammernik • D. Rückert

Links

URL

Research Area

 C1 | Medicine

BibTeXKey: ZSP+24b

Back to Top