{"id":{"repo_id":"usf","oai_identifier":"oai:digitalcommons.usf.edu:etd-11862"},"canonical_url":"https://search.ndltd.org/etd/usf/oai:digitalcommons.usf.edu:etd-11862","repository":{"repo_id":"usf","name":"University of South Florida","base_url":"https://digitalcommons.usf.edu/do/oai/"},"display":{"title":"From Graph Theory for Robust Deep Networks to Graph Learning for Multimodal Cancer Analysis","abstract":"This dissertation explores the intersection of graph theory and deep learning, focusing on enhancing the robustness of deep neural networks (DNNs) and applying these advancements to complex problems like cancer diagnosis and treatment. We investigate the structural properties of graphs and their influence on neural network performance, particularly in multimodal learning. The work delves into the design space of DNN architectures using graph-theoretic measures, transforming graphs into DNN architectures for various tasks, and examining their robustness against noise and adversarial attacks. The study extends to medical imaging, highlighting advanced DNN architectures like U-Net for brain tumor segmentation. It addresses the evolution of digital pathology, the challenges of task-specific AI/ML models, and the transformative potential of foundation models and generative AI. The integration of multimodal oncology data through Graph Neural Networks (GNNs) and Transformers is explored, showcasing their potential in improving diagnostic and prognostic models. The development of the Multimodal Integration of Oncology Data System (MINDS) and SeNMo, a deep learning model for multi-omics data, underscores the significance of harmonizing diverse data types for personalized cancer care. We also proposed a GNN-based hierarchical relational model, PARADIGM, that enhances survival predictions by integrating multimodal datasets. The compilation of articles is structured into eight chapters, each focusing on different aspects of learning, from theoretical foundations to practical applications, offering a comprehensive overview of the field and its implications for future research and clinical practice in oncology and computational pathology.","abstract_html":"This dissertation explores the intersection of graph theory and deep learning, focusing on enhancing the robustness of deep neural networks (DNNs) and applying these advancements to complex problems like cancer diagnosis and treatment. We investigate the structural properties of graphs and their influence on neural network performance, particularly in multimodal learning. The work delves into the design space of DNN architectures using graph-theoretic measures, transforming graphs into DNN architectures for various tasks, and examining their robustness against noise and adversarial attacks. The study extends to medical imaging, highlighting advanced DNN architectures like U-Net for brain tumor segmentation. It addresses the evolution of digital pathology, the challenges of task-specific AI/ML models, and the transformative potential of foundation models and generative AI. The integration of multimodal oncology data through Graph Neural Networks (GNNs) and Transformers is explored, showcasing their potential in improving diagnostic and prognostic models. The development of the Multimodal Integration of Oncology Data System (MINDS) and SeNMo, a deep learning model for multi-omics data, underscores the significance of harmonizing diverse data types for personalized cancer care. We also proposed a GNN-based hierarchical relational model, PARADIGM, that enhances survival predictions by integrating multimodal datasets. The compilation of articles is structured into eight chapters, each focusing on different aspects of learning, from theoretical foundations to practical applications, offering a comprehensive overview of the field and its implications for future research and clinical practice in oncology and computational pathology.","abstract_has_math":false,"creators":["Waqas, Asim"],"institution":"University of South Florida","degree_name":"Doctor of Philosophy (Ph.D.)","degree_level":"Dissertation","degree_discipline":"Electrical Engineering","degree_department":null,"school":null,"contributors":["Ghulam Rasool, Ph.D.","Mia Naeini, Ph.D.","Nidhal Carla Bouaynaya, Ph.D.","Ismail Uysal, Ph.D.","Paul Stewart, Ph.D."],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2024,"date_issued":"2024-06-25T07:00:00Z","date_published":"2024-06-25T07:00:00Z","updated_at":"2026-11-19T08:50:14Z","subjects":["Deep Learning","Foundation Models","Graph","Neural Architecture Search","Pan-cancer","Artificial Intelligence and Robotics","Computer Sciences","Oncology"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://digitalcommons.usf.edu/etd/10575","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://digitalcommons.usf.edu/etd/10575","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://digitalcommons.usf.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Adigitalcommons.usf.edu%3Aetd-11862","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Ghulam Rasool, Ph.D.","Mia Naeini, Ph.D.","Nidhal Carla Bouaynaya, Ph.D.","Ismail Uysal, Ph.D.","Paul Stewart, Ph.D."]},{"key":"dc:creator","label":"Author","values":["Waqas, Asim"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2024-06-25T07:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation","Ph.D."]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy (Ph.D.)"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of South Florida"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Deep Learning","Foundation Models","Graph","Neural Architecture Search","Pan-cancer","Artificial Intelligence and Robotics","Computer Sciences","Oncology"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://digitalcommons.usf.edu/etd/10575"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["This dissertation explores the intersection of graph theory and deep learning, focusing on enhancing the robustness of deep neural networks (DNNs) and applying these advancements to complex problems like cancer diagnosis and treatment. We investigate the structural properties of graphs and their influence on neural network performance, particularly in multimodal learning. The work delves into the design space of DNN architectures using graph-theoretic measures, transforming graphs into DNN architectures for various tasks, and examining their robustness against noise and adversarial attacks. The study extends to medical imaging, highlighting advanced DNN architectures like U-Net for brain tumor segmentation. It addresses the evolution of digital pathology, the challenges of task-specific AI/ML models, and the transformative potential of foundation models and generative AI. The integration of multimodal oncology data through Graph Neural Networks (GNNs) and Transformers is explored, showcasing their potential in improving diagnostic and prognostic models. The development of the Multimodal Integration of Oncology Data System (MINDS) and SeNMo, a deep learning model for multi-omics data, underscores the significance of harmonizing diverse data types for personalized cancer care. We also proposed a GNN-based hierarchical relational model, PARADIGM, that enhances survival predictions by integrating multimodal datasets. The compilation of articles is structured into eight chapters, each focusing on different aspects of learning, from theoretical foundations to practical applications, offering a comprehensive overview of the field and its implications for future research and clinical practice in oncology and computational pathology."]},{"key":"dc:title","label":"Title","values":["From Graph Theory for Robust Deep Networks to Graph Learning for Multimodal Cancer Analysis"]}]}],"canonical_facts":{"dc:contributor":["Ghulam Rasool, Ph.D.","Mia Naeini, Ph.D.","Nidhal Carla Bouaynaya, Ph.D.","Ismail Uysal, Ph.D.","Paul Stewart, Ph.D."],"dc:creator":["Waqas, Asim"],"dc:date":["2024-06-25T07:00:00Z"],"dc:description":["This dissertation explores the intersection of graph theory and deep learning, focusing on enhancing the robustness of deep neural networks (DNNs) and applying these advancements to complex problems like cancer diagnosis and treatment. We investigate the structural properties of graphs and their influence on neural network performance, particularly in multimodal learning. The work delves into the design space of DNN architectures using graph-theoretic measures, transforming graphs into DNN architectures for various tasks, and examining their robustness against noise and adversarial attacks. The study extends to medical imaging, highlighting advanced DNN architectures like U-Net for brain tumor segmentation. It addresses the evolution of digital pathology, the challenges of task-specific AI/ML models, and the transformative potential of foundation models and generative AI. The integration of multimodal oncology data through Graph Neural Networks (GNNs) and Transformers is explored, showcasing their potential in improving diagnostic and prognostic models. The development of the Multimodal Integration of Oncology Data System (MINDS) and SeNMo, a deep learning model for multi-omics data, underscores the significance of harmonizing diverse data types for personalized cancer care. We also proposed a GNN-based hierarchical relational model, PARADIGM, that enhances survival predictions by integrating multimodal datasets. The compilation of articles is structured into eight chapters, each focusing on different aspects of learning, from theoretical foundations to practical applications, offering a comprehensive overview of the field and its implications for future research and clinical practice in oncology and computational pathology."],"dc:identifier":["https://digitalcommons.usf.edu/etd/10575"],"dc:subject":["Deep Learning","Foundation Models","Graph","Neural Architecture Search","Pan-cancer","Artificial Intelligence and Robotics","Computer Sciences","Oncology"],"dc:title":["From Graph Theory for Robust Deep Networks to Graph Learning for Multimodal Cancer Analysis"],"thesis:degree_discipline":["Electrical Engineering"],"thesis:degree_level":["Dissertation","Ph.D."],"thesis:degree_name":["Doctor of Philosophy (Ph.D.)"],"thesis:institution_name":["University of South Florida"]},"updated_at":"2026-11-19T08:50:14Z"}