Unsupervised MRI reconstruction via zero-shot learned adversarial transformers

Simple item page
buir.contributor.authorKorkmaz, Yilmaz
buir.contributor.authorDar, Salman U.H.
buir.contributor.authorYurt, Mahmut
buir.contributor.authorÖzbey, Muzaffer
buir.contributor.authorÇukur, Tolga
buir.contributor.orcidÇukur, Tolga|0000-0002-2296-851X
dc.citation.epage1763en_US
dc.citation.issueNumber7en_US
dc.citation.spage1747en_US
dc.citation.volumeNumber41en_US
dc.contributor.authorKorkmaz, Yilmaz
dc.contributor.authorDar, Salman U.H.
dc.contributor.authorYurt, Mahmut
dc.contributor.authorÖzbey, Muzaffer
dc.contributor.authorÇukur, Tolga
dc.date.accessioned2023-02-15T12:17:49Z
dc.date.available2023-02-15T12:17:49Z
dc.date.issued2022-01-27
dc.departmentDepartment of Electrical and Electronics Engineeringen_US
dc.description.abstractSupervised reconstruction models are characteristically trained on matched pairs of undersampled and fully-sampled data to capture an MRI prior, along with supervision regarding the imaging operator to enforce data consistency. To reduce supervision requirements, the recent deep image prior framework instead conjoins untrained MRI priors with the imaging operator during inference. Yet, canonical convolutional architectures are suboptimal in capturing long-range relationships, and priors based on randomly initialized networks may yield suboptimal performance. To address these limitations, here we introduce a novel unsupervised MRI reconstruction method based on zero-Shot Learned Adversarial TransformERs (SLATER). SLATER embodies a deep adversarial network with cross-attention transformers to map noise and latent variables onto coil-combined MR images. During pre-training, this unconditional network learns a high-quality MRI prior in an unsupervised generative modeling task. During inference, a zero-shot reconstruction is then performed by incorporating the imaging operator and optimizing the prior to maximize consistency to undersampled data. Comprehensive experiments on brain MRI datasets clearly demonstrate the superior performance of SLATER against state-of-the-art unsupervised methods.en_US
dc.description.provenanceSubmitted by Fatma Kaya (fattttoky.55@gmail.com) on 2023-02-15T12:17:49Z No. of bitstreams: 1 Unsupervised_MRI_reconstruction_via_zero-shot_learned_adversarial_transformers.pdf: 17559252 bytes, checksum: 8e88af13c7ae4d4fd289b2e5c18e5b22 (MD5)en
dc.description.provenanceMade available in DSpace on 2023-02-15T12:17:49Z (GMT). No. of bitstreams: 1 Unsupervised_MRI_reconstruction_via_zero-shot_learned_adversarial_transformers.pdf: 17559252 bytes, checksum: 8e88af13c7ae4d4fd289b2e5c18e5b22 (MD5) Previous issue date: 2022-01-27en
dc.identifier.doi10.1109/TMI.2022.3147426en_US
dc.identifier.eissn1558-254X
dc.identifier.issn0278-0062
dc.identifier.urihttp://hdl.handle.net/11693/111353
dc.language.isoEnglishen_US
dc.publisherInstitute of Electrical and Electronics Engineers Inc.en_US
dc.relation.isversionofhttps://www.doi.org/10.1109/TMI.2022.3147426en_US
dc.source.titleIEEE Transactions on Medical Imagingen_US
dc.subjectAdversarialen_US
dc.subjectTransformersen_US
dc.subjectMRIen_US
dc.subjectUnsuperviseden_US
dc.subjectReconstructionen_US
dc.subjectZero shoten_US
dc.subjectGenerativeen_US
dc.titleUnsupervised MRI reconstruction via zero-shot learned adversarial transformersen_US
dc.typeArticleen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Unsupervised_MRI_reconstruction_via_zero-shot_learned_adversarial_transformers.pdf
Size:
16.75 MB
Format:
Adobe Portable Document Format
Description:
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.69 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

Scholarly Publications - Electrical and Electronics Engineering