{"id":{"repo_id":"buffalo","oai_identifier":"oai:ubir.buffalo.edu:10477/84045"},"canonical_url":"https://search.dev.ndltd.org/etd/buffalo/oai:ubir.buffalo.edu:10477/84045","repository":{"repo_id":"buffalo","name":"Buffalo","base_url":"https://ubir.buffalo.edu/oai/request"},"display":{"title":"Exploring the Utility of Bayesian Networks in Histopathological Image Analysis: Beyond Classifier Networks","abstract":"M.S.","abstract_html":"M.S.","abstract_has_math":false,"creators":["Border, Samuel; 0000-0002-8806-6965"],"institution":"State University of New York at Buffalo","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Sarder, Pinaki","Biomedical Engineering"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-06-21T15:47:21Z","date_published":"2022-06-21T15:47:21Z","updated_at":"2026-07-27T19:05:30Z","subjects":["artificial intelligence","biomedical engineering","medical imaging"],"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/10477/84045","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Sarder, Pinaki","Biomedical Engineering"]},{"key":"dc:creator","label":"Author","values":["Border, Samuel; 0000-0002-8806-6965"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2022-06-21T15:47:21Z","2020"]},{"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":["artificial intelligence","biomedical engineering","medical imaging"]}]},{"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/10477/84045"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["M.S.","Pathologists' assessment of tissue histology is primarily confined to the study select spatial features of interest for making diagnostic and/or prognostic decision. In recent years, research scientists have applied high-throughput computational methods such as neural networks (NN) capable of learning their own descriptive features to minimize prediction cost. For image related projects convolutional neural networks have proven highly effective and have surged in popularity over a variety of tasks. However, a major drawback of NN tools is the significant gap in understanding between researchers and clinicians. In a clinical environment it is important to establish the biological relevance of the feature sets that are important for informed decision making. Bayesian networks (BNs), a type of probabilistic graphical model, have not been widely adapted in computational pathology despite the wide variety of potential benefits. They allow clinicians to incorporate hand-crafted features from a wide range of modalities to assess disease progression on a continuous scale. In this thesis, we have explored the different functionalities of BNs and how they can be utilized to better understand the factors contributing to progression of Diabetic Nephropathy (DN). Our key results include DN stage classification with absolute error (AE) of 1.25 using a BN constructed from n = 1124 human DN glomeruli. This result outperforms classical machine learning methods, such as k-nearest neighbor, decision trees, and SVM. Although more modern classification methods (MLP, AdaBoost, Nearest Centroid) achieved a slightly lower AE, the benefit of BN is the ability to directly model feature relationships to impart more information on underlying disease mechanisms.","**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":["Exploring the Utility of Bayesian Networks in Histopathological Image Analysis: Beyond Classifier Networks"]}]}],"canonical_facts":{"dc:contributor":["Sarder, Pinaki","Biomedical Engineering"],"dc:creator":["Border, Samuel; 0000-0002-8806-6965"],"dc:date":["2022-06-21T15:47:21Z","2020"],"dc:description":["M.S.","Pathologists' assessment of tissue histology is primarily confined to the study select spatial features of interest for making diagnostic and/or prognostic decision. In recent years, research scientists have applied high-throughput computational methods such as neural networks (NN) capable of learning their own descriptive features to minimize prediction cost. For image related projects convolutional neural networks have proven highly effective and have surged in popularity over a variety of tasks. However, a major drawback of NN tools is the significant gap in understanding between researchers and clinicians. In a clinical environment it is important to establish the biological relevance of the feature sets that are important for informed decision making. Bayesian networks (BNs), a type of probabilistic graphical model, have not been widely adapted in computational pathology despite the wide variety of potential benefits. They allow clinicians to incorporate hand-crafted features from a wide range of modalities to assess disease progression on a continuous scale. In this thesis, we have explored the different functionalities of BNs and how they can be utilized to better understand the factors contributing to progression of Diabetic Nephropathy (DN). Our key results include DN stage classification with absolute error (AE) of 1.25 using a BN constructed from n = 1124 human DN glomeruli. This result outperforms classical machine learning methods, such as k-nearest neighbor, decision trees, and SVM. Although more modern classification methods (MLP, AdaBoost, Nearest Centroid) achieved a slightly lower AE, the benefit of BN is the ability to directly model feature relationships to impart more information on underlying disease mechanisms.","**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/10477/84045"],"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":["artificial intelligence","biomedical engineering","medical imaging"],"dc:title":["Exploring the Utility of Bayesian Networks in Histopathological Image Analysis: Beyond Classifier Networks"],"dc:type":["Text","Thesis"]},"updated_at":"2026-07-27T19:05:30Z"}