{"id":{"repo_id":"temple","oai_identifier":"oai:scholarshare.temple.edu:20.500.12613/4073"},"canonical_url":"https://search.dev.ndltd.org/etd/temple/oai:scholarshare.temple.edu:20.500.12613/4073","repository":{"repo_id":"temple","name":"Temple University","base_url":"https://scholarshare.temple.edu/server/oai/request"},"display":{"title":"POCS Augmented CycleGAN for MR Image Reconstruction","abstract":"Traditional Magnetic Resonance Imaging (MRI) reconstruction methods, which may be highly time-consuming and sensitive to noise, heavily depend on solving nonlinear optimization problems. By contrast, deep learning (DL)-based reconstruction methods do not need any explicit analytical data model and are robust to noise due to its large data-based training, which both make DL a versatile tool for fast and high-fidelity MR image reconstruction. While DL can be performed completely independently of traditional methods, it can, in fact, benefit from incorporating these established methods to achieve better results. To test this hypothesis, we proposed a hybrid DL-based MR image reconstruction method, which combines two state-of-the-art deep learning networks, U-Net and Generative Adversarial Network with Cycle loss (CycleGAN), with a traditional data reconstruction method: Projection Onto Convex Sets (POCS). Experiments were then performed to evaluate the method by comparing it to several existing state-of-the-art methods. Our results demonstrate that the proposed method outperformed the current state-of-the-art in terms of higher peak signal-to-noise ratio (PSNR) and higher Structural Similarity Index (SSIM).","abstract_html":"Traditional Magnetic Resonance Imaging (MRI) reconstruction methods, which may be highly time-consuming and sensitive to noise, heavily depend on solving nonlinear optimization problems. By contrast, deep learning (DL)-based reconstruction methods do not need any explicit analytical data model and are robust to noise due to its large data-based training, which both make DL a versatile tool for fast and high-fidelity MR image reconstruction. While DL can be performed completely independently of traditional methods, it can, in fact, benefit from incorporating these established methods to achieve better results. To test this hypothesis, we proposed a hybrid DL-based MR image reconstruction method, which combines two state-of-the-art deep learning networks, U-Net and Generative Adversarial Network with Cycle loss (CycleGAN), with a traditional data reconstruction method: Projection Onto Convex Sets (POCS). Experiments were then performed to evaluate the method by comparing it to several existing state-of-the-art methods. Our results demonstrate that the proposed method outperformed the current state-of-the-art in terms of higher peak signal-to-noise ratio (PSNR) and higher Structural Similarity Index (SSIM).","abstract_has_math":false,"creators":["Yang, Hanlu"],"institution":"Temple University. Libraries","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":["Bai, Li"],"committee_chairs":[],"committee_members":["Bai, Li","Wang, Ze","Ahmad, Fauzia (Electrical engineer)"],"year":2020,"date_issued":"2020","date_published":"2020","updated_at":"2026-07-27T21:20:51Z","subjects":["Electrical engineering","Compressed sensing","Cyclegan","Deep learning","MR image reconstruction"],"languages":["eng"],"rights":["IN COPYRIGHT- This Rights Statement can be used for an Item that is in copyright. Using this statement implies that the organization making this Item available has determined that the Item is in copyright and either is the rights-holder, has obtained permission from the rights-holder(s) to make their Work(s) available, or makes the Item available under an exception or limitation to copyright (including Fair Use) that entitles it to make the Item available."],"rights_urls":["http://rightsstatements.org/vocab/InC/1.0/"],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/20.500.12613/4073","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Bai, Li"]},{"key":"dc:contributor.committeemember","label":"Committee Member","values":["Bai, Li","Wang, Ze","Ahmad, Fauzia (Electrical engineer)"]},{"key":"dc:creator","label":"Author","values":["Yang, Hanlu"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2020-11-05T19:50:44Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2020-11-05T19:50:44Z"]},{"key":"dc:date.issued","label":"Date","values":["2020"]},{"key":"dc:publisher","label":"Institution","values":["Temple University. Libraries"]},{"key":"dc:type","label":"Dc Type","values":["Text"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Electrical engineering","Compressed sensing","Cyclegan","Deep learning","MR image reconstruction"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["IN COPYRIGHT- This Rights Statement can be used for an Item that is in copyright. Using this statement implies that the organization making this Item available has determined that the Item is in copyright and either is the rights-holder, has obtained permission from the rights-holder(s) to make their Work(s) available, or makes the Item available under an exception or limitation to copyright (including Fair Use) that entitles it to make the Item available."]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://rightsstatements.org/vocab/InC/1.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["http://hdl.handle.net/20.500.12613/4073"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Traditional Magnetic Resonance Imaging (MRI) reconstruction methods, which may be highly time-consuming and sensitive to noise, heavily depend on solving nonlinear optimization problems. By contrast, deep learning (DL)-based reconstruction methods do not need any explicit analytical data model and are robust to noise due to its large data-based training, which both make DL a versatile tool for fast and high-fidelity MR image reconstruction. While DL can be performed completely independently of traditional methods, it can, in fact, benefit from incorporating these established methods to achieve better results. To test this hypothesis, we proposed a hybrid DL-based MR image reconstruction method, which combines two state-of-the-art deep learning networks, U-Net and Generative Adversarial Network with Cycle loss (CycleGAN), with a traditional data reconstruction method: Projection Onto Convex Sets (POCS). Experiments were then performed to evaluate the method by comparing it to several existing state-of-the-art methods. Our results demonstrate that the proposed method outperformed the current state-of-the-art in terms of higher peak signal-to-noise ratio (PSNR) and higher Structural Similarity Index (SSIM)."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["M.S.E.E."]},{"key":"dc:title","label":"Title","values":["POCS Augmented CycleGAN for MR Image Reconstruction"]}]}],"canonical_facts":{"dc:contributor.advisor":["Bai, Li"],"dc:contributor.committeemember":["Bai, Li","Wang, Ze","Ahmad, Fauzia (Electrical engineer)"],"dc:creator":["Yang, Hanlu"],"dc:date.accessioned":["2020-11-05T19:50:44Z"],"dc:date.available":["2020-11-05T19:50:44Z"],"dc:date.issued":["2020"],"dc:description.abstract":["Traditional Magnetic Resonance Imaging (MRI) reconstruction methods, which may be highly time-consuming and sensitive to noise, heavily depend on solving nonlinear optimization problems. By contrast, deep learning (DL)-based reconstruction methods do not need any explicit analytical data model and are robust to noise due to its large data-based training, which both make DL a versatile tool for fast and high-fidelity MR image reconstruction. While DL can be performed completely independently of traditional methods, it can, in fact, benefit from incorporating these established methods to achieve better results. To test this hypothesis, we proposed a hybrid DL-based MR image reconstruction method, which combines two state-of-the-art deep learning networks, U-Net and Generative Adversarial Network with Cycle loss (CycleGAN), with a traditional data reconstruction method: Projection Onto Convex Sets (POCS). Experiments were then performed to evaluate the method by comparing it to several existing state-of-the-art methods. Our results demonstrate that the proposed method outperformed the current state-of-the-art in terms of higher peak signal-to-noise ratio (PSNR) and higher Structural Similarity Index (SSIM)."],"dc:description.degree":["M.S.E.E."],"dc:identifier.uri":["http://hdl.handle.net/20.500.12613/4073"],"dc:language.iso":["eng"],"dc:publisher":["Temple University. Libraries"],"dc:rights":["IN COPYRIGHT- This Rights Statement can be used for an Item that is in copyright. Using this statement implies that the organization making this Item available has determined that the Item is in copyright and either is the rights-holder, has obtained permission from the rights-holder(s) to make their Work(s) available, or makes the Item available under an exception or limitation to copyright (including Fair Use) that entitles it to make the Item available."],"dc:rights.uri":["http://rightsstatements.org/vocab/InC/1.0/"],"dc:subject":["Electrical engineering","Compressed sensing","Cyclegan","Deep learning","MR image reconstruction"],"dc:title":["POCS Augmented CycleGAN for MR Image Reconstruction"],"dc:type":["Text"]},"updated_at":"2026-07-27T21:20:51Z"}