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

Download:

Current browse context:

eess.SP

Change to browse by:

References & Citations

Bookmark

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

Electrical Engineering and Systems Science > Signal Processing

Title: Deep J-Sense: Accelerated MRI Reconstruction via Unrolled Alternating Optimization

Abstract: Accelerated multi-coil magnetic resonance imaging reconstruction has seen a substantial recent improvement combining compressed sensing with deep learning. However, most of these methods rely on estimates of the coil sensitivity profiles, or on calibration data for estimating model parameters. Prior work has shown that these methods degrade in performance when the quality of these estimators are poor or when the scan parameters differ from the training conditions. Here we introduce Deep J-Sense as a deep learning approach that builds on unrolled alternating minimization and increases robustness: our algorithm refines both the magnetization (image) kernel and the coil sensitivity maps. Experimental results on a subset of the knee fastMRI dataset show that this increases reconstruction performance and provides a significant degree of robustness to varying acceleration factors and calibration region sizes.
Subjects: Signal Processing (eess.SP); Machine Learning (cs.LG)
Cite as: arXiv:2103.02087 [eess.SP]
  (or arXiv:2103.02087v3 [eess.SP] for this version)

Submission history

From: Marius Arvinte [view email]
[v1] Tue, 2 Mar 2021 23:22:22 GMT (8157kb,D)
[v2] Thu, 1 Apr 2021 14:35:37 GMT (7526kb,D)
[v3] Sun, 11 Apr 2021 17:02:47 GMT (7524kb,D)

Link back to: arXiv, form interface, contact.