We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Image and Video Processing

Title: WKGM: Weight-K-space Generative Model for Parallel Imaging Reconstruction

Abstract: Deep learning based parallel imaging (PI) has made great progresses in recent years to accelerate magnetic resonance imaging (MRI). Nevertheless, it still has some limitations, such as the robustness and flexibility of existing methods have great deficiency. In this work, we propose a method to explore the k-space domain learning via robust generative modeling for flexible calibration-less PI reconstruction, coined weight-k-space generative model (WKGM). Specifically, WKGM is a generalized k-space domain model, where the k-space weighting technology and high-dimensional space augmentation design are efficiently incorporated for score-based generative model training, resulting in good and robust reconstructions. In addition, WKGM is flexible and thus can be synergistically combined with various traditional k-space PI models, which can make full use of the correlation between multi-coil data and realizecalibration-less PI. Even though our model was trained on only 500 images, experimental results with varying sampling patterns and acceleration factors demonstrate that WKGM can attain state-of-the-art reconstruction results with the well-learned k-space generative prior.
Comments: 11pages, 12 figures
Subjects: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2205.03883 [eess.IV]
  (or arXiv:2205.03883v4 [eess.IV] for this version)

Submission history

From: Qiegen Liu [view email]
[v1] Sun, 8 May 2022 14:28:20 GMT (1768kb)
[v2] Tue, 14 Jun 2022 04:25:42 GMT (1846kb)
[v3] Mon, 21 Nov 2022 02:01:30 GMT (1841kb)
[v4] Fri, 25 Nov 2022 01:58:40 GMT (1841kb)

Link back to: arXiv, form interface, contact.