DC Field | Value | Language |
---|---|---|
dc.contributor.author | Oh, Changheun | ko |
dc.contributor.author | Kim, Dongchan | ko |
dc.contributor.author | Chung, Jun-Young | ko |
dc.contributor.author | Han, Yeji | ko |
dc.contributor.author | Park, HyunWook | ko |
dc.date.accessioned | 2021-03-04T04:50:05Z | - |
dc.date.available | 2021-03-04T04:50:05Z | - |
dc.date.created | 2020-12-28 | - |
dc.date.issued | 2021-01 | - |
dc.identifier.citation | Medical Physics, v.48, no.1, pp.193 - 203 | - |
dc.identifier.issn | 0094-2405 | - |
dc.identifier.uri | http://hdl.handle.net/10203/281182 | - |
dc.description.abstract | Purpose Reconstructing the images from undersampled k-space data are an ill-posed inverse problem. As a solution to this problem, we propose a method to reconstruct magnetic resonance (MR) images directly from k-space data using a recurrent neural network. Methods A novel neural network architecture named "ETER-net" is developed as a unified solution to reconstruct MR images from undersampled k-space data, where two bi-RNNs and convolutional neural network (CNN) are utilized to perform domain transformation and de-aliasing. To demonstrate the practicality of the proposed method, we conducted model optimization, cross-validation, and network pruning using in-house data from a 3T MRI scanner and public dataset called "FastMRI." Results The experimental results showed that the proposed method could be utilized for accurate image reconstruction from undersampled k-space data. The size of the proposed model was optimized and cross-validation was performed to show the robustness of the proposed method. For in-house dataset (R = 4), the proposed method provided nMSE = 1.09% and SSIM = 0.938. For "FastMRI" dataset, the proposed method provided nMSE = 1.05 % and SSIM = 0.931 for R = 4, and nMSE = 3.12 % and SSIM = 0.884 for R = 8. The performance of the pruned model trained the loss function including with L2 regularization was consistent for a pruning ratio of up to 70%. Conclusions The proposed method is an end-to-end MR image reconstruction method based on recurrent neural networks. It performs direct mapping of the input k-space data and the reconstructed images, operating as a unified solution that is applicable to various scanning trajectories. | - |
dc.language | English | - |
dc.publisher | American Association of Physicists in Medicine | - |
dc.title | A k-space-to-image reconstruction network for MRI using recurrent neural network | - |
dc.type | Article | - |
dc.identifier.wosid | 000597760900001 | - |
dc.identifier.scopusid | 2-s2.0-85097429690 | - |
dc.type.rims | ART | - |
dc.citation.volume | 48 | - |
dc.citation.issue | 1 | - |
dc.citation.beginningpage | 193 | - |
dc.citation.endingpage | 203 | - |
dc.citation.publicationname | Medical Physics | - |
dc.identifier.doi | 10.1002/mp.14566 | - |
dc.contributor.localauthor | Park, HyunWook | - |
dc.contributor.nonIdAuthor | Kim, Dongchan | - |
dc.contributor.nonIdAuthor | Chung, Jun-Young | - |
dc.contributor.nonIdAuthor | Han, Yeji | - |
dc.description.isOpenAccess | N | - |
dc.type.journalArticle | Article | - |
dc.subject.keywordAuthor | deep learning | - |
dc.subject.keywordAuthor | end& | - |
dc.subject.keywordAuthor | #8208 | - |
dc.subject.keywordAuthor | to& | - |
dc.subject.keywordAuthor | #8208 | - |
dc.subject.keywordAuthor | end reconstruction network (ETER& | - |
dc.subject.keywordAuthor | #8208 | - |
dc.subject.keywordAuthor | net) | - |
dc.subject.keywordAuthor | MR image reconstruction | - |
dc.subject.keywordAuthor | parallel imaging | - |
dc.subject.keywordAuthor | recurrent neural network | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.