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

Download:

Current browse context:

physics

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 > Image and Video Processing

Title: Single-pass Object-adaptive Data Undersampling and Reconstruction for MRI

Abstract: There is much recent interest in techniques to accelerate the data acquisition process in MRI by acquiring limited measurements. Often sophisticated reconstruction algorithms are deployed to maintain high image quality in such settings. In this work, we propose a data-driven sampler using a convolutional neural network, MNet, to provide object-specific sampling patterns adaptive to each scanned object. The network observes very limited low-frequency k-space data for each object and rapidly predicts the desired undersampling pattern in one go that achieves high image reconstruction quality. We propose an accompanying alternating-type training framework with a mask-backward procedure that efficiently generates training labels for the sampler network and jointly trains an image reconstruction network. Experimental results on the fastMRI knee dataset demonstrate the ability of the proposed learned undersampling network to generate object-specific masks at fourfold and eightfold acceleration that achieve superior image reconstruction performance than several existing schemes. The source code for the proposed joint sampling and reconstruction learning framework is available at this https URL
Subjects: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Medical Physics (physics.med-ph)
Journal reference: in IEEE Transactions on Computational Imaging, vol. 8, pp. 333-345, 2022
DOI: 10.1109/TCI.2022.3167454
Cite as: arXiv:2111.09212 [eess.IV]
  (or arXiv:2111.09212v3 [eess.IV] for this version)

Submission history

From: Zhishen Huang [view email]
[v1] Wed, 17 Nov 2021 16:06:06 GMT (4838kb,D)
[v2] Wed, 27 Apr 2022 16:03:21 GMT (9454kb,D)
[v3] Wed, 18 May 2022 20:53:42 GMT (9449kb,D)

Link back to: arXiv, form interface, contact.