{"id":{"repo_id":"cambridge","oai_identifier":"oai:www.repository.cam.ac.uk:1810/387076"},"canonical_url":"https://search.dev.ndltd.org/etd/cambridge/oai:www.repository.cam.ac.uk:1810/387076","repository":{"repo_id":"cambridge","name":"Cambridge University","base_url":"https://api.repository.cam.ac.uk/server/oai/request"},"display":{"title":"The role of image processing and artificial intelligence machine learning-based systems for unbiased evaluation of human coronary atherosclerotic plaque progression and vulnerability, and in response to pharmacotherapy","abstract":"Intracoronary optical coherence tomography (OCT) can identify high-risk plaques and is a widely used surrogate efficacy marker for drug and device studies. For example, fibrous cap thickness (FCT) <75µm, minimum lumen area (MLA) <3.5mm2, lipid arc >180°, and presence of macrophages, calcific nodules, neovascularisation, and cholesterol crystals are associated with major adverse coronary events (MACE). Many of these features change with drug or device therapy, including drugs that reduce patient events with minimal changes in plaque volume. However, real-world OCT pullbacks are rich datasets containing hundreds of images and tens-of-thousands of candidate measurements per artery. Consequently, detailed OCT analysis requires time-consuming offline manual frame selection and measurement in specialised core laboratories. Furthermore, inter- and intra-observer variability for particular tissues is suboptimal, even between core laboratories. Analysis is also limited by the high frequency of artifacts, and similarity of artifact to disease. The work presented in this thesis first demonstrates the burden of artifacts with a large range of imaging artifacts present in OCT pullbacks. Pre-processing with a novel artifact correction method improved sensitivity and diagnostic accuracy to detect fibrous plaques and fibroatheroma. Second, this work also attempted to mitigate the subjectivity of OCT interpretation by quantifying accepted descriptions demonstrating that parameters such as peak pixel intensity and lesion edge gradient are discriminative of normal tissue and lipid tissue, respectively, but the diagnostic ability of these parameters is poor. Therefore, a deep learning AI-based image analysis system for intracoronary OCT (AutoOCT) was designed and tested. It was able to detect and measure multiple markers of coronary artery disease, and used pre-processing to mitigate effects of artifacts, optimise poor quality images, and allow analysis of all available data. AutoOCT measurements showed high accuracy, identifying features of drug efficacy and changes in plaque morphology in response to high-intensity statins. It also showed non-inferior diagnostic performance to detect vulnerable plaque features in an automated analysis of CLIMA (Relationship Between Coronary Plaque Morphology of Left Anterior Descending Artery and Long-Term Clinical Outcome study). While AI-based OCT analysis may not replace human interpretation, whole vessel and frame-based analysis may greatly speed up the process and reduce intra- and interobserver variability.","abstract_html":"Intracoronary optical coherence tomography (OCT) can identify high-risk plaques and is a widely used surrogate efficacy marker for drug and device studies. For example, fibrous cap thickness (FCT) &lt;75µm, minimum lumen area (MLA) &lt;3.5mm2, lipid arc &gt;180°, and presence of macrophages, calcific nodules, neovascularisation, and cholesterol crystals are associated with major adverse coronary events (MACE). Many of these features change with drug or device therapy, including drugs that reduce patient events with minimal changes in plaque volume. However, real-world OCT pullbacks are rich datasets containing hundreds of images and tens-of-thousands of candidate measurements per artery. Consequently, detailed OCT analysis requires time-consuming offline manual frame selection and measurement in specialised core laboratories. Furthermore, inter- and intra-observer variability for particular tissues is suboptimal, even between core laboratories. Analysis is also limited by the high frequency of artifacts, and similarity of artifact to disease. The work presented in this thesis first demonstrates the burden of artifacts with a large range of imaging artifacts present in OCT pullbacks. Pre-processing with a novel artifact correction method improved sensitivity and diagnostic accuracy to detect fibrous plaques and fibroatheroma. Second, this work also attempted to mitigate the subjectivity of OCT interpretation by quantifying accepted descriptions demonstrating that parameters such as peak pixel intensity and lesion edge gradient are discriminative of normal tissue and lipid tissue, respectively, but the diagnostic ability of these parameters is poor. Therefore, a deep learning AI-based image analysis system for intracoronary OCT (AutoOCT) was designed and tested. It was able to detect and measure multiple markers of coronary artery disease, and used pre-processing to mitigate effects of artifacts, optimise poor quality images, and allow analysis of all available data. AutoOCT measurements showed high accuracy, identifying features of drug efficacy and changes in plaque morphology in response to high-intensity statins. It also showed non-inferior diagnostic performance to detect vulnerable plaque features in an automated analysis of CLIMA (Relationship Between Coronary Plaque Morphology of Left Anterior Descending Artery and Long-Term Clinical Outcome study). While AI-based OCT analysis may not replace human interpretation, whole vessel and frame-based analysis may greatly speed up the process and reduce intra- and interobserver variability.","abstract_has_math":false,"creators":["Jessney, Benn"],"institution":"University of Cambridge","degree_name":"Doctor of Philosophy (PhD)","degree_level":"Doctoral","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":["Bennett, Martin"],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-01-07","date_published":"2025-01-07","updated_at":"2026-07-22T22:24:21Z","subjects":["Artefact","Artificial Intelligence","Assistive Diagnostics","Atherosclerosis","Coronary Artery Disease","Image Processing","Invasive Imaging","Machine Learning","Optical Coherence Tomography","Vulnerable Plaque"],"languages":[],"rights":[],"rights_urls":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/54b4a127-3aa9-4164-8654-741876500911/download","http://purl.org/NET/rdflicense/allrightsreserved"],"identifier_entries":[]},"links":{"outbound_url":"https://doi.org/10.17863/CAM.120008","outbound_label":"DOI","outbound_source":"dc:identifier.doi"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Bennett, Martin"]},{"key":"dc:contributor.sponsor","label":"Sponsor","values":["Supported by British Heart Foundation Grants PG/18/14/33562, RG13/14/30314, RE/24/130011, TA/F/20/210001 (London), Academy of Medical Sciences Starter Grants for Clinical Lecturers (REF: SGL030\\1012), Innovate UK Advancing Precision Medicine 10069871, National Institutes of Health, R01 HL150608, EPSRC Cambridge Maths in Healthcare (Nr. EP/N014588/1, and Cambridge NIHR Biomedical Research Centres."]},{"key":"dc:creator","label":"Author","values":["Jessney, Benn"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.issued","label":"Date","values":["2025-01-07"]},{"key":"dc:publisher.institution","label":"Dc Publisher Institution","values":["University of Cambridge"]},{"key":"dc:relation.isreferencedby.uri","label":"Dc Relation Isreferencedby URI","values":["https://www.repository.cam.ac.uk/handle/1810/387076"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"dc:type.qualificationlevel","label":"Dc Type Qualificationlevel","values":["Doctoral"]},{"key":"dc:type.qualificationname","label":"Dc Type Qualificationname","values":["Doctor of Philosophy (PhD)"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Artefact","Artificial Intelligence","Assistive Diagnostics","Atherosclerosis","Coronary Artery Disease","Image Processing","Invasive Imaging","Machine Learning","Optical Coherence Tomography","Vulnerable Plaque"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/54b4a127-3aa9-4164-8654-741876500911/download","http://purl.org/NET/rdflicense/allrightsreserved"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.doi","label":"DOI","values":["https://doi.org/10.17863/CAM.120008"]},{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/14c02aa2-4a50-4e75-8264-6d1d8c75574f/download"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Intracoronary optical coherence tomography (OCT) can identify high-risk plaques and is a widely used surrogate efficacy marker for drug and device studies. For example, fibrous cap thickness (FCT) <75µm, minimum lumen area (MLA) <3.5mm2, lipid arc >180°, and presence of macrophages, calcific nodules, neovascularisation, and cholesterol crystals are associated with major adverse coronary events (MACE). Many of these features change with drug or device therapy, including drugs that reduce patient events with minimal changes in plaque volume. However, real-world OCT pullbacks are rich datasets containing hundreds of images and tens-of-thousands of candidate measurements per artery. Consequently, detailed OCT analysis requires time-consuming offline manual frame selection and measurement in specialised core laboratories. Furthermore, inter- and intra-observer variability for particular tissues is suboptimal, even between core laboratories. Analysis is also limited by the high frequency of artifacts, and similarity of artifact to disease. The work presented in this thesis first demonstrates the burden of artifacts with a large range of imaging artifacts present in OCT pullbacks. Pre-processing with a novel artifact correction method improved sensitivity and diagnostic accuracy to detect fibrous plaques and fibroatheroma. Second, this work also attempted to mitigate the subjectivity of OCT interpretation by quantifying accepted descriptions demonstrating that parameters such as peak pixel intensity and lesion edge gradient are discriminative of normal tissue and lipid tissue, respectively, but the diagnostic ability of these parameters is poor. Therefore, a deep learning AI-based image analysis system for intracoronary OCT (AutoOCT) was designed and tested. It was able to detect and measure multiple markers of coronary artery disease, and used pre-processing to mitigate effects of artifacts, optimise poor quality images, and allow analysis of all available data. AutoOCT measurements showed high accuracy, identifying features of drug efficacy and changes in plaque morphology in response to high-intensity statins. It also showed non-inferior diagnostic performance to detect vulnerable plaque features in an automated analysis of CLIMA (Relationship Between Coronary Plaque Morphology of Left Anterior Descending Artery and Long-Term Clinical Outcome study). While AI-based OCT analysis may not replace human interpretation, whole vessel and frame-based analysis may greatly speed up the process and reduce intra- and interobserver variability."]},{"key":"dc:format.checksum.md5","label":"Dc Format Checksum Md5","values":["45523cf4a77871bdb1990eca11cf23d0","87eda9de84448d1f82354d60eee3eb5f"]},{"key":"dc:title","label":"Title","values":["The role of image processing and artificial intelligence machine learning-based systems for unbiased evaluation of human coronary atherosclerotic plaque progression and vulnerability, and in response to pharmacotherapy"]}]}],"canonical_facts":{"dc:contributor.advisor":["Bennett, Martin"],"dc:contributor.sponsor":["Supported by British Heart Foundation Grants PG/18/14/33562, RG13/14/30314, RE/24/130011, TA/F/20/210001 (London), Academy of Medical Sciences Starter Grants for Clinical Lecturers (REF: SGL030\\1012), Innovate UK Advancing Precision Medicine 10069871, National Institutes of Health, R01 HL150608, EPSRC Cambridge Maths in Healthcare (Nr. EP/N014588/1, and Cambridge NIHR Biomedical Research Centres."],"dc:creator":["Jessney, Benn"],"dc:date.issued":["2025-01-07"],"dc:description.abstract":["Intracoronary optical coherence tomography (OCT) can identify high-risk plaques and is a widely used surrogate efficacy marker for drug and device studies. For example, fibrous cap thickness (FCT) <75µm, minimum lumen area (MLA) <3.5mm2, lipid arc >180°, and presence of macrophages, calcific nodules, neovascularisation, and cholesterol crystals are associated with major adverse coronary events (MACE). Many of these features change with drug or device therapy, including drugs that reduce patient events with minimal changes in plaque volume. However, real-world OCT pullbacks are rich datasets containing hundreds of images and tens-of-thousands of candidate measurements per artery. Consequently, detailed OCT analysis requires time-consuming offline manual frame selection and measurement in specialised core laboratories. Furthermore, inter- and intra-observer variability for particular tissues is suboptimal, even between core laboratories. Analysis is also limited by the high frequency of artifacts, and similarity of artifact to disease. The work presented in this thesis first demonstrates the burden of artifacts with a large range of imaging artifacts present in OCT pullbacks. Pre-processing with a novel artifact correction method improved sensitivity and diagnostic accuracy to detect fibrous plaques and fibroatheroma. Second, this work also attempted to mitigate the subjectivity of OCT interpretation by quantifying accepted descriptions demonstrating that parameters such as peak pixel intensity and lesion edge gradient are discriminative of normal tissue and lipid tissue, respectively, but the diagnostic ability of these parameters is poor. Therefore, a deep learning AI-based image analysis system for intracoronary OCT (AutoOCT) was designed and tested. It was able to detect and measure multiple markers of coronary artery disease, and used pre-processing to mitigate effects of artifacts, optimise poor quality images, and allow analysis of all available data. AutoOCT measurements showed high accuracy, identifying features of drug efficacy and changes in plaque morphology in response to high-intensity statins. It also showed non-inferior diagnostic performance to detect vulnerable plaque features in an automated analysis of CLIMA (Relationship Between Coronary Plaque Morphology of Left Anterior Descending Artery and Long-Term Clinical Outcome study). While AI-based OCT analysis may not replace human interpretation, whole vessel and frame-based analysis may greatly speed up the process and reduce intra- and interobserver variability."],"dc:format.checksum.md5":["45523cf4a77871bdb1990eca11cf23d0","87eda9de84448d1f82354d60eee3eb5f"],"dc:identifier.doi":["https://doi.org/10.17863/CAM.120008"],"dc:identifier.uri":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/14c02aa2-4a50-4e75-8264-6d1d8c75574f/download"],"dc:publisher.institution":["University of Cambridge"],"dc:relation.isreferencedby.uri":["https://www.repository.cam.ac.uk/handle/1810/387076"],"dc:rights":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/54b4a127-3aa9-4164-8654-741876500911/download","http://purl.org/NET/rdflicense/allrightsreserved"],"dc:subject":["Artefact","Artificial Intelligence","Assistive Diagnostics","Atherosclerosis","Coronary Artery Disease","Image Processing","Invasive Imaging","Machine Learning","Optical Coherence Tomography","Vulnerable Plaque"],"dc:title":["The role of image processing and artificial intelligence machine learning-based systems for unbiased evaluation of human coronary atherosclerotic plaque progression and vulnerability, and in response to pharmacotherapy"],"dc:type":["Thesis"],"dc:type.qualificationlevel":["Doctoral"],"dc:type.qualificationname":["Doctor of Philosophy (PhD)"]},"updated_at":"2026-07-22T22:24:21Z"}