{"id":{"repo_id":"buffalo","oai_identifier":"oai:ubir.buffalo.edu:10477/86822"},"canonical_url":"https://search.dev.ndltd.org/etd/buffalo/oai:ubir.buffalo.edu:10477/86822","repository":{"repo_id":"buffalo","name":"Buffalo","base_url":"https://ubir.buffalo.edu/oai/request"},"display":{"title":"Multiple GAN Augmentation for Acute Myocardial Infarction Classification","abstract":"M.Eng.","abstract_html":"M.Eng.","abstract_has_math":false,"creators":["Zhao, Kanhao"],"institution":"State University of New York at Buffalo","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Leslie, Ying","Biomedical Engineering"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-02-25T23:23:07Z","date_published":"2025-02-25T23:23:07Z","updated_at":"2026-07-27T19:05:37Z","subjects":["biomedical engineering"],"languages":["eng"],"rights":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/10477/86822","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Leslie, Ying","Biomedical Engineering"]},{"key":"dc:creator","label":"Author","values":["Zhao, Kanhao"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2025-02-25T23:23:07Z","2020","2020-07-27 21:19:16"]},{"key":"dc:publisher","label":"Institution","values":["State University of New York at Buffalo"]},{"key":"dc:type","label":"Dc Type","values":["Text","Thesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["biomedical engineering"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/10477/86822"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["M.Eng.","Convolutional Neural Networks (CNNs) achieve excellent outcome from various given image tasks with large labeled training data. However, medical images are hard to collected and therefore, most medical datasets are small. In this thesis, generating diverse and reliable images by Generative Adversarial Networks (GANs) are proposed as a new augmentation method, which can introduce additional information to original dataset and improve CNN classification accuracy. GAN variants are widely designed specially in two scenarios: 1) images generated from random vectors, and 2) translated images from source domain to target domain. Progressive growing of GAN (PGGAN) and Cycle-consistent GAN (CylceGAN) are representative of GAN from each scenario. We compare the GAN-based augmentation results with these two GANs. The CNN backbone in our research is MobilenetV2[15], and channel attention module is introduced to enhance the CNN performance in advance. The best classification accuracy of 92.7% training from PGGAN augmentation demonstrates that GAN-based data augmentation can be effective method to assist CNN classification.","**To request an accessible version of the file(s) associated with this item, contact library@buffalo.edu. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.**"]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Multiple GAN Augmentation for Acute Myocardial Infarction Classification"]}]}],"canonical_facts":{"dc:contributor":["Leslie, Ying","Biomedical Engineering"],"dc:creator":["Zhao, Kanhao"],"dc:date":["2025-02-25T23:23:07Z","2020","2020-07-27 21:19:16"],"dc:description":["M.Eng.","Convolutional Neural Networks (CNNs) achieve excellent outcome from various given image tasks with large labeled training data. However, medical images are hard to collected and therefore, most medical datasets are small. In this thesis, generating diverse and reliable images by Generative Adversarial Networks (GANs) are proposed as a new augmentation method, which can introduce additional information to original dataset and improve CNN classification accuracy. GAN variants are widely designed specially in two scenarios: 1) images generated from random vectors, and 2) translated images from source domain to target domain. Progressive growing of GAN (PGGAN) and Cycle-consistent GAN (CylceGAN) are representative of GAN from each scenario. We compare the GAN-based augmentation results with these two GANs. The CNN backbone in our research is MobilenetV2[15], and channel attention module is introduced to enhance the CNN performance in advance. The best classification accuracy of 92.7% training from PGGAN augmentation demonstrates that GAN-based data augmentation can be effective method to assist CNN classification.","**To request an accessible version of the file(s) associated with this item, contact library@buffalo.edu. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.**"],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/10477/86822"],"dc:language":["eng"],"dc:publisher":["State University of New York at Buffalo"],"dc:rights":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."],"dc:subject":["biomedical engineering"],"dc:title":["Multiple GAN Augmentation for Acute Myocardial Infarction Classification"],"dc:type":["Text","Thesis"]},"updated_at":"2026-07-27T19:05:37Z"}