File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: MRI reconstruction using deep Bayesian estimation

TitleMRI reconstruction using deep Bayesian estimation
Authors
KeywordsBayesian estimation
compressed sensing
deep learning reconstruction
generative network
parallel imaging
Issue Date2020
PublisherJohn Wiley & Sons, Inc. The Journal's web site is located at http://www.interscience.wiley.com/jpages/0740-3194/
Citation
Magnetic Resonance in Medicine, 2020, Epub 2020-04-10 How to Cite?
AbstractPurpose: To develop a deep learning‐based Bayesian estimation for MRI reconstruction. Methods: We modeled the MRI reconstruction problem with Bayes’s theorem, following the recently proposed PixelCNN++ method. The image reconstruction from incomplete k‐space measurement was obtained by maximizing the posterior possibility. A generative network was utilized as the image prior, which was computationally tractable, and the k‐space data fidelity was enforced by using an equality constraint. The stochastic backpropagation was utilized to calculate the descent gradient in the process of maximum a posterior, and a projected subgradient method was used to impose the equality constraint. In contrast to the other deep learning reconstruction methods, the proposed one used the likelihood of prior as the training loss and the objective function in reconstruction to improve the image quality. Results The proposed method showed an improved performance in preserving image details and reducing aliasing artifacts, compared with GRAPPA, urn:x-wiley:07403194:media:mrm28274:mrm28274-math-0003‐ESPRiT, model‐based deep learning architecture for inverse problems (MODL), and variational network (VN), last two were state‐of‐the‐art deep learning reconstruction methods. The proposed method generally achieved more than 3 dB peak signal‐to‐noise ratio improvement for compressed sensing and parallel imaging reconstructions compared with the other methods. Conclusions The Bayesian estimation significantly improved the reconstruction performance, compared with the conventional urn:x-wiley:07403194:media:mrm28274:mrm28274-math-0004‐sparsity prior in compressed sensing reconstruction tasks. More importantly, the proposed reconstruction framework can be generalized for most MRI reconstruction scenarios.
Persistent Identifierhttp://hdl.handle.net/10722/282911
ISSN
2019 Impact Factor: 3.635
2015 SCImago Journal Rankings: 2.197
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorLUO, G-
dc.contributor.authorZHAO, N-
dc.contributor.authorJiang, W-
dc.contributor.authorHui, ES-
dc.contributor.authorCao, P-
dc.date.accessioned2020-06-05T06:22:59Z-
dc.date.available2020-06-05T06:22:59Z-
dc.date.issued2020-
dc.identifier.citationMagnetic Resonance in Medicine, 2020, Epub 2020-04-10-
dc.identifier.issn0740-3194-
dc.identifier.urihttp://hdl.handle.net/10722/282911-
dc.description.abstractPurpose: To develop a deep learning‐based Bayesian estimation for MRI reconstruction. Methods: We modeled the MRI reconstruction problem with Bayes’s theorem, following the recently proposed PixelCNN++ method. The image reconstruction from incomplete k‐space measurement was obtained by maximizing the posterior possibility. A generative network was utilized as the image prior, which was computationally tractable, and the k‐space data fidelity was enforced by using an equality constraint. The stochastic backpropagation was utilized to calculate the descent gradient in the process of maximum a posterior, and a projected subgradient method was used to impose the equality constraint. In contrast to the other deep learning reconstruction methods, the proposed one used the likelihood of prior as the training loss and the objective function in reconstruction to improve the image quality. Results The proposed method showed an improved performance in preserving image details and reducing aliasing artifacts, compared with GRAPPA, urn:x-wiley:07403194:media:mrm28274:mrm28274-math-0003‐ESPRiT, model‐based deep learning architecture for inverse problems (MODL), and variational network (VN), last two were state‐of‐the‐art deep learning reconstruction methods. The proposed method generally achieved more than 3 dB peak signal‐to‐noise ratio improvement for compressed sensing and parallel imaging reconstructions compared with the other methods. Conclusions The Bayesian estimation significantly improved the reconstruction performance, compared with the conventional urn:x-wiley:07403194:media:mrm28274:mrm28274-math-0004‐sparsity prior in compressed sensing reconstruction tasks. More importantly, the proposed reconstruction framework can be generalized for most MRI reconstruction scenarios.-
dc.languageeng-
dc.publisherJohn Wiley & Sons, Inc. The Journal's web site is located at http://www.interscience.wiley.com/jpages/0740-3194/-
dc.relation.ispartofMagnetic Resonance in Medicine-
dc.rightsPreprint This is the pre-peer reviewed version of the following article: [FULL CITE], which has been published in final form at [Link to final article using the DOI]. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions. Postprint This is the peer reviewed version of the following article: [FULL CITE], which has been published in final form at [Link to final article using the DOI]. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions.-
dc.subjectBayesian estimation-
dc.subjectcompressed sensing-
dc.subjectdeep learning reconstruction-
dc.subjectgenerative network-
dc.subjectparallel imaging-
dc.titleMRI reconstruction using deep Bayesian estimation-
dc.typeArticle-
dc.identifier.emailHui, ES: edshui@hku.hk-
dc.identifier.emailCao, P: caopeng1@hku.hk-
dc.identifier.authorityHui, ES=rp01832-
dc.identifier.authorityCao, P=rp02474-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1002/mrm.28274-
dc.identifier.scopuseid_2-s2.0-85083064578-
dc.identifier.hkuros310277-
dc.identifier.volumeEpub 2020-04-10-
dc.identifier.isiWOS:000559831200001-
dc.publisher.placeUnited States-
dc.identifier.issnl0740-3194-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats