{"id":{"repo_id":"mtu","oai_identifier":"oai:digitalcommons.mtu.edu:etdr-2700"},"canonical_url":"https://search.ndltd.org/etd/mtu/oai:digitalcommons.mtu.edu:etdr-2700","repository":{"repo_id":"mtu","name":"Michigan Technological University","base_url":"https://digitalcommons.mtu.edu/do/oai/"},"display":{"title":"Artificial Intelligence Algorithms for Eye Banking","abstract":"Eye banking plays a critical role in modern medicine by providing cornea tissues for transplantation to restore vision for millions of people worldwide. The evaluation of corneal endothelium is done by measuring the corneal endothelial cell density (ECD). Unfortunately, the current system to measure ECD is manual, time-consuming, and error prone. Furthermore, the impact of social behaviors and biological conditions on corneal endothelium and corneal transplant success is largely unexplored. To overcome these challenges, this dissertation aims to develop tools for corneal endothelial image and data analysis that enhance the efficiency and quality of the cornea transplants. In the first study, an image processing algorithm is developed to analyze corneal endothelial images captured by a Konan CellChek specular microscope. The algorithm successfully identifies the region of interest, filters the image, and employs stochastic watershed segmentation to determine cell boundaries and evaluate endothelial cell density (ECD). The proposed algorithm achieves a high correlation with manual counts (R2 = 0.98) and has an average analysis time of 2.5 seconds. In the second study, a deep learning-based cell segmentation algorithm called Mobile-CellNet is proposed to estimate ECD. This technique addresses the limitations of classical algorithms and creates a more robust and highly efficient algorithm. The approach achieves a mean absolute error of 4.06% for ECD on the test set, similar to U-Net but with significantly fewer floating-point operations and parameters. The third study explores the correlation between alcohol abuse and corneal endothelial morphology in a donor pool of 5,624 individuals. Multivariable regression analysis shows that alcohol abuse is associated with a reduction in endothelial cell density, an increase in the coefficient of variation, and a decrease in percent hexagonality. These studies highlight the potential of big data and artificial algorithms in accurately and efficiently analyzing corneal images and donor medical data to improve the efficiency of eye banking and patient outcomes. By automating the analysis of corneal images and exploring the impact of social behaviors and biological conditions on corneal endothelial morphology, we can enhance the quality and availability of cornea transplants and ultimately improve the lives of millions of people worldwide.","abstract_html":"Eye banking plays a critical role in modern medicine by providing cornea tissues for transplantation to restore vision for millions of people worldwide. The evaluation of corneal endothelium is done by measuring the corneal endothelial cell density (ECD). Unfortunately, the current system to measure ECD is manual, time-consuming, and error prone. Furthermore, the impact of social behaviors and biological conditions on corneal endothelium and corneal transplant success is largely unexplored. To overcome these challenges, this dissertation aims to develop tools for corneal endothelial image and data analysis that enhance the efficiency and quality of the cornea transplants. In the first study, an image processing algorithm is developed to analyze corneal endothelial images captured by a Konan CellChek specular microscope. The algorithm successfully identifies the region of interest, filters the image, and employs stochastic watershed segmentation to determine cell boundaries and evaluate endothelial cell density (ECD). The proposed algorithm achieves a high correlation with manual counts (R2 = 0.98) and has an average analysis time of 2.5 seconds. In the second study, a deep learning-based cell segmentation algorithm called Mobile-CellNet is proposed to estimate ECD. This technique addresses the limitations of classical algorithms and creates a more robust and highly efficient algorithm. The approach achieves a mean absolute error of 4.06% for ECD on the test set, similar to U-Net but with significantly fewer floating-point operations and parameters. The third study explores the correlation between alcohol abuse and corneal endothelial morphology in a donor pool of 5,624 individuals. Multivariable regression analysis shows that alcohol abuse is associated with a reduction in endothelial cell density, an increase in the coefficient of variation, and a decrease in percent hexagonality. These studies highlight the potential of big data and artificial algorithms in accurately and efficiently analyzing corneal images and donor medical data to improve the efficiency of eye banking and patient outcomes. By automating the analysis of corneal images and exploring the impact of social behaviors and biological conditions on corneal endothelial morphology, we can enhance the quality and availability of cornea transplants and ultimately improve the lives of millions of people worldwide.","abstract_has_math":false,"creators":["Karmakar, Ranit"],"institution":null,"degree_name":"Doctor of Philosophy in Computer Engineering (PhD)","degree_level":"Open Access Dissertation","degree_discipline":"Department of Electrical and Computer Engineering","degree_department":null,"school":null,"contributors":["Sean Kirkpatrick"],"advisors":[],"committee_chairs":[],"committee_members":["Glen Archer","Daniel Fuhrmann","Allen Eghrari"],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-11-19T08:18:46Z","subjects":["eye banking; cornea; artificial intelligence; segmentation; data analysis;"],"languages":[],"rights":[],"rights_urls":["http://creativecommons.org/licenses/by-nc/4.0/"],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["https://digitalcommons.mtu.edu/etdr/1545"],"render_values":[{"text":"https://digitalcommons.mtu.edu/etdr/1545","href":"https://digitalcommons.mtu.edu/etdr/1545","code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://doi.org/10.37099/mtu.dc.etdr/1545","outbound_label":"DOI","outbound_source":"dc:identifier","record_url":"https://doi.org/10.37099/mtu.dc.etdr/1545","record_label":"Repository record","record_kind":"doi","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.mtu.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Adigitalcommons.mtu.edu%3Aetdr-2700","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Sean Kirkpatrick"]},{"key":"dc:contributor.committeemember","label":"Committee Member","values":["Glen Archer","Daniel Fuhrmann","Allen Eghrari"]},{"key":"dc:creator","label":"Author","values":["Karmakar, Ranit"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.created","label":"Dc Date Created","values":["2023-01-01T08:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Department of Electrical and Computer Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Open Access Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy in Computer Engineering (PhD)"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["eye banking; cornea; artificial intelligence; segmentation; data analysis;"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights.license","label":"Dc Rights License","values":["http://creativecommons.org/licenses/by-nc/4.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["10.37099/mtu.dc.etdr/1545","https://digitalcommons.mtu.edu/etdr/1545"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Eye banking plays a critical role in modern medicine by providing cornea tissues for transplantation to restore vision for millions of people worldwide. The evaluation of corneal endothelium is done by measuring the corneal endothelial cell density (ECD). Unfortunately, the current system to measure ECD is manual, time-consuming, and error prone. Furthermore, the impact of social behaviors and biological conditions on corneal endothelium and corneal transplant success is largely unexplored. To overcome these challenges, this dissertation aims to develop tools for corneal endothelial image and data analysis that enhance the efficiency and quality of the cornea transplants. In the first study, an image processing algorithm is developed to analyze corneal endothelial images captured by a Konan CellChek specular microscope. The algorithm successfully identifies the region of interest, filters the image, and employs stochastic watershed segmentation to determine cell boundaries and evaluate endothelial cell density (ECD). The proposed algorithm achieves a high correlation with manual counts (R2 = 0.98) and has an average analysis time of 2.5 seconds. In the second study, a deep learning-based cell segmentation algorithm called Mobile-CellNet is proposed to estimate ECD. This technique addresses the limitations of classical algorithms and creates a more robust and highly efficient algorithm. The approach achieves a mean absolute error of 4.06% for ECD on the test set, similar to U-Net but with significantly fewer floating-point operations and parameters. The third study explores the correlation between alcohol abuse and corneal endothelial morphology in a donor pool of 5,624 individuals. Multivariable regression analysis shows that alcohol abuse is associated with a reduction in endothelial cell density, an increase in the coefficient of variation, and a decrease in percent hexagonality. These studies highlight the potential of big data and artificial algorithms in accurately and efficiently analyzing corneal images and donor medical data to improve the efficiency of eye banking and patient outcomes. By automating the analysis of corneal images and exploring the impact of social behaviors and biological conditions on corneal endothelial morphology, we can enhance the quality and availability of cornea transplants and ultimately improve the lives of millions of people worldwide."]},{"key":"dc:title","label":"Title","values":["Artificial Intelligence Algorithms for Eye Banking"]}]}],"canonical_facts":{"dc:contributor":["Sean Kirkpatrick"],"dc:contributor.committeemember":["Glen Archer","Daniel Fuhrmann","Allen Eghrari"],"dc:creator":["Karmakar, Ranit"],"dc:date.created":["2023-01-01T08:00:00Z"],"dc:description.abstract":["Eye banking plays a critical role in modern medicine by providing cornea tissues for transplantation to restore vision for millions of people worldwide. The evaluation of corneal endothelium is done by measuring the corneal endothelial cell density (ECD). Unfortunately, the current system to measure ECD is manual, time-consuming, and error prone. Furthermore, the impact of social behaviors and biological conditions on corneal endothelium and corneal transplant success is largely unexplored. To overcome these challenges, this dissertation aims to develop tools for corneal endothelial image and data analysis that enhance the efficiency and quality of the cornea transplants. In the first study, an image processing algorithm is developed to analyze corneal endothelial images captured by a Konan CellChek specular microscope. The algorithm successfully identifies the region of interest, filters the image, and employs stochastic watershed segmentation to determine cell boundaries and evaluate endothelial cell density (ECD). The proposed algorithm achieves a high correlation with manual counts (R2 = 0.98) and has an average analysis time of 2.5 seconds. In the second study, a deep learning-based cell segmentation algorithm called Mobile-CellNet is proposed to estimate ECD. This technique addresses the limitations of classical algorithms and creates a more robust and highly efficient algorithm. The approach achieves a mean absolute error of 4.06% for ECD on the test set, similar to U-Net but with significantly fewer floating-point operations and parameters. The third study explores the correlation between alcohol abuse and corneal endothelial morphology in a donor pool of 5,624 individuals. Multivariable regression analysis shows that alcohol abuse is associated with a reduction in endothelial cell density, an increase in the coefficient of variation, and a decrease in percent hexagonality. These studies highlight the potential of big data and artificial algorithms in accurately and efficiently analyzing corneal images and donor medical data to improve the efficiency of eye banking and patient outcomes. By automating the analysis of corneal images and exploring the impact of social behaviors and biological conditions on corneal endothelial morphology, we can enhance the quality and availability of cornea transplants and ultimately improve the lives of millions of people worldwide."],"dc:identifier":["10.37099/mtu.dc.etdr/1545","https://digitalcommons.mtu.edu/etdr/1545"],"dc:rights.license":["http://creativecommons.org/licenses/by-nc/4.0/"],"dc:subject":["eye banking; cornea; artificial intelligence; segmentation; data analysis;"],"dc:title":["Artificial Intelligence Algorithms for Eye Banking"],"thesis:degree_discipline":["Department of Electrical and Computer Engineering"],"thesis:degree_level":["Open Access Dissertation"],"thesis:degree_name":["Doctor of Philosophy in Computer Engineering (PhD)"]},"updated_at":"2026-11-19T08:18:46Z"}