{"id":{"repo_id":"qu-belfast","oai_identifier":"oai:pure.qub.ac.uk/portal:studenttheses/449e1757-6776-45cb-b2a3-96c4606a5651"},"canonical_url":"https://search.dev.ndltd.org/etd/qu-belfast/oai:pure.qub.ac.uk/portal:studenttheses/449e1757-6776-45cb-b2a3-96c4606a5651","repository":{"repo_id":"qu-belfast","name":"Queen's University Belfast","base_url":"https://pureadmin.qub.ac.uk/ws/oai"},"display":{"title":"Detection and tracking of cells and focal adhesions in microscopy images","abstract":"Significant progress in the biomedical field (including cancer research, immunology and drug discovery) has been fuelled by an improved understanding of cell function and structure. Modern microscopy and imaging techniques have enabled cell biologists to observe and record cell processes in detail, thus generating high volumes of imaging data. Meaningful conclusions can only be drawn from such big datasets using automated quantitative analysis. <br/><br/>This thesis focuses on ways to automatically perform detection, segmentation, and tracking of cells or subcellular structures in time-lapse microscopy image sequences to facilitate efficient analysis of these datasets. Furthermore, a secondary purpose of the thesis deals with the lack of annotated data in the biomedical field. <br/><br/>In order to achieve these goals, we propose the use of probabilistic tracking algorithms to accurately and robustly track either migrating cells or their internal structures. For example, focal adhesions moving inside a single migrating normal human epidermal keratinocyte. In particular, the use of the Interacting Multiple Model (IMM) algorithm is proposed given its ability to take into account several different motion models at every given time. Within this framework, we formulate the data association problem as two linear assignment problems and we experiment with first and second order interacting models, different state vector definitions and hard or soft estimation. Regarding detection and segmentation, we propose several convolutional neural network architectures to detect and segment cells. The proposed networks have the advantage of being able to learn with very limited and incomplete training data. Moreover, we present a U-Net based multi-task learning architecture for simultaneously segmenting and detecting cells. <br/><br/>In summary, the main contributions of this thesis are a multitarget tracking framework for identical targets, a novel integrated framework for focal adhesion detection and tracking, a U-Net based cell detector, and a novel multi-task U-Net architecture for simultaneously localising and segmenting cells in phase-contrast microscopy images.","abstract_html":"Significant progress in the biomedical field (including cancer research, immunology and drug discovery) has been fuelled by an improved understanding of cell function and structure. Modern microscopy and imaging techniques have enabled cell biologists to observe and record cell processes in detail, thus generating high volumes of imaging data. Meaningful conclusions can only be drawn from such big datasets using automated quantitative analysis. &lt;br/&gt;&lt;br/&gt;This thesis focuses on ways to automatically perform detection, segmentation, and tracking of cells or subcellular structures in time-lapse microscopy image sequences to facilitate efficient analysis of these datasets. Furthermore, a secondary purpose of the thesis deals with the lack of annotated data in the biomedical field. &lt;br/&gt;&lt;br/&gt;In order to achieve these goals, we propose the use of probabilistic tracking algorithms to accurately and robustly track either migrating cells or their internal structures. For example, focal adhesions moving inside a single migrating normal human epidermal keratinocyte. In particular, the use of the Interacting Multiple Model (IMM) algorithm is proposed given its ability to take into account several different motion models at every given time. Within this framework, we formulate the data association problem as two linear assignment problems and we experiment with first and second order interacting models, different state vector definitions and hard or soft estimation. Regarding detection and segmentation, we propose several convolutional neural network architectures to detect and segment cells. The proposed networks have the advantage of being able to learn with very limited and incomplete training data. Moreover, we present a U-Net based multi-task learning architecture for simultaneously segmenting and detecting cells. &lt;br/&gt;&lt;br/&gt;In summary, the main contributions of this thesis are a multitarget tracking framework for identical targets, a novel integrated framework for focal adhesion detection and tracking, a U-Net based cell detector, and a novel multi-task U-Net architecture for simultaneously localising and segmenting cells in phase-contrast microscopy images.","abstract_has_math":false,"creators":["Lomanov, Katerina"],"institution":"Queen's University Belfast","degree_name":"Doctor of Philosophy","degree_level":"Doctoral Thesis","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":["Miller, Paul","Martinez del Rincon, Jesus"],"committee_chairs":[],"committee_members":[],"year":2020,"date_issued":"2020-12","date_published":"2020-12","updated_at":"2026-07-24T03:55:31Z","subjects":["Tracking","Detection"],"languages":["eng"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["oai:pure.qub.ac.uk/portal:studenttheses/449e1757-6776-45cb-b2a3-96c4606a5651"],"render_values":[{"text":"oai:pure.qub.ac.uk/portal:studenttheses/449e1757-6776-45cb-b2a3-96c4606a5651","href":null,"code":true}]}]},"links":{"outbound_url":"https://pure.qub.ac.uk/en/studentTheses/449e1757-6776-45cb-b2a3-96c4606a5651","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Miller, Paul","Martinez del Rincon, Jesus"]},{"key":"dc:contributor.sponsor","label":"Sponsor","values":["EC/Horizon 2020 Marie Skłodowska-Curie actions"]},{"key":"dc:creator","label":"Author","values":["Lomanov, Katerina"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2020-12"]},{"key":"dc:date.issued","label":"Date","values":["2020-12"]},{"key":"dc:publisher.department","label":"Dc Publisher Department","values":["School of Electronics, Electrical Engineering and Computer Science"]},{"key":"dc:publisher.institution","label":"Dc Publisher Institution","values":["Queen's University Belfast"]},{"key":"dc:relation.isreferencedby","label":"Dc Relation Isreferencedby","values":["https://pure.qub.ac.uk/en/studentTheses/449e1757-6776-45cb-b2a3-96c4606a5651"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"dc:type.qualificationlevel","label":"Dc Type Qualificationlevel","values":["Doctoral Thesis"]},{"key":"dc:type.qualificationname","label":"Dc Type Qualificationname","values":["Doctor of Philosophy"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Tracking","Detection"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["oai:pure.qub.ac.uk/portal:studenttheses/449e1757-6776-45cb-b2a3-96c4606a5651","https://pure.qub.ac.uk/en/studentTheses/449e1757-6776-45cb-b2a3-96c4606a5651"]},{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://pure.qub.ac.uk/files/216121025/thesis_Katerina.pdf"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Significant progress in the biomedical field (including cancer research, immunology and drug discovery) has been fuelled by an improved understanding of cell function and structure. Modern microscopy and imaging techniques have enabled cell biologists to observe and record cell processes in detail, thus generating high volumes of imaging data. Meaningful conclusions can only be drawn from such big datasets using automated quantitative analysis. <br/><br/>This thesis focuses on ways to automatically perform detection, segmentation, and tracking of cells or subcellular structures in time-lapse microscopy image sequences to facilitate efficient analysis of these datasets. Furthermore, a secondary purpose of the thesis deals with the lack of annotated data in the biomedical field. <br/><br/>In order to achieve these goals, we propose the use of probabilistic tracking algorithms to accurately and robustly track either migrating cells or their internal structures. For example, focal adhesions moving inside a single migrating normal human epidermal keratinocyte. In particular, the use of the Interacting Multiple Model (IMM) algorithm is proposed given its ability to take into account several different motion models at every given time. Within this framework, we formulate the data association problem as two linear assignment problems and we experiment with first and second order interacting models, different state vector definitions and hard or soft estimation. Regarding detection and segmentation, we propose several convolutional neural network architectures to detect and segment cells. The proposed networks have the advantage of being able to learn with very limited and incomplete training data. Moreover, we present a U-Net based multi-task learning architecture for simultaneously segmenting and detecting cells. <br/><br/>In summary, the main contributions of this thesis are a multitarget tracking framework for identical targets, a novel integrated framework for focal adhesion detection and tracking, a U-Net based cell detector, and a novel multi-task U-Net architecture for simultaneously localising and segmenting cells in phase-contrast microscopy images."]},{"key":"dc:title","label":"Title","values":["Detection and tracking of cells and focal adhesions in microscopy images"]}]}],"canonical_facts":{"dc:contributor.advisor":["Miller, Paul","Martinez del Rincon, Jesus"],"dc:contributor.sponsor":["EC/Horizon 2020 Marie Skłodowska-Curie actions"],"dc:creator":["Lomanov, Katerina"],"dc:date":["2020-12"],"dc:date.issued":["2020-12"],"dc:description.abstract":["Significant progress in the biomedical field (including cancer research, immunology and drug discovery) has been fuelled by an improved understanding of cell function and structure. Modern microscopy and imaging techniques have enabled cell biologists to observe and record cell processes in detail, thus generating high volumes of imaging data. Meaningful conclusions can only be drawn from such big datasets using automated quantitative analysis. <br/><br/>This thesis focuses on ways to automatically perform detection, segmentation, and tracking of cells or subcellular structures in time-lapse microscopy image sequences to facilitate efficient analysis of these datasets. Furthermore, a secondary purpose of the thesis deals with the lack of annotated data in the biomedical field. <br/><br/>In order to achieve these goals, we propose the use of probabilistic tracking algorithms to accurately and robustly track either migrating cells or their internal structures. For example, focal adhesions moving inside a single migrating normal human epidermal keratinocyte. In particular, the use of the Interacting Multiple Model (IMM) algorithm is proposed given its ability to take into account several different motion models at every given time. Within this framework, we formulate the data association problem as two linear assignment problems and we experiment with first and second order interacting models, different state vector definitions and hard or soft estimation. Regarding detection and segmentation, we propose several convolutional neural network architectures to detect and segment cells. The proposed networks have the advantage of being able to learn with very limited and incomplete training data. Moreover, we present a U-Net based multi-task learning architecture for simultaneously segmenting and detecting cells. <br/><br/>In summary, the main contributions of this thesis are a multitarget tracking framework for identical targets, a novel integrated framework for focal adhesion detection and tracking, a U-Net based cell detector, and a novel multi-task U-Net architecture for simultaneously localising and segmenting cells in phase-contrast microscopy images."],"dc:identifier":["oai:pure.qub.ac.uk/portal:studenttheses/449e1757-6776-45cb-b2a3-96c4606a5651","https://pure.qub.ac.uk/en/studentTheses/449e1757-6776-45cb-b2a3-96c4606a5651"],"dc:identifier.uri":["https://pure.qub.ac.uk/files/216121025/thesis_Katerina.pdf"],"dc:language":["eng"],"dc:publisher.department":["School of Electronics, Electrical Engineering and Computer Science"],"dc:publisher.institution":["Queen's University Belfast"],"dc:relation.isreferencedby":["https://pure.qub.ac.uk/en/studentTheses/449e1757-6776-45cb-b2a3-96c4606a5651"],"dc:subject":["Tracking","Detection"],"dc:title":["Detection and tracking of cells and focal adhesions in microscopy images"],"dc:type":["Thesis"],"dc:type.qualificationlevel":["Doctoral Thesis"],"dc:type.qualificationname":["Doctor of Philosophy"]},"updated_at":"2026-07-24T03:55:31Z"}