{"id":{"repo_id":"cambridge","oai_identifier":"oai:www.repository.cam.ac.uk:1810/342480"},"canonical_url":"https://search.dev.ndltd.org/etd/cambridge/oai:www.repository.cam.ac.uk:1810/342480","repository":{"repo_id":"cambridge","name":"Cambridge University","base_url":"https://api.repository.cam.ac.uk/server/oai/request"},"display":{"title":"Evaluating Artificial Intelligence in Breast Cancer Screening","abstract":"This thesis evaluates the application and performance of artificial intelligence (AI) in breast cancer screening. Breast cancer screening is conducted on a population scale using mammographic imaging for the earlier detection of breast cancer and has been shown to reduce mortality. A shortage of trained radiologists, as well as the demands of double reading, mean an approach to alleviate pressures within the breast screening workflow is sought. In addition, interval cancers occur at an estimated rate of 3.7/1000 women screened in the UK, thus methods to improve the sensitivity of screening and detect cancers earlier are also needed. Advances in AI over the past decade have demonstrated comparable performance to human readers and could provide a method for an adapted screening workflow to improve both efficiency and efficacy of screening. However, the 2021 National Screening Committee (NSC) report concluded that there was insufficient evidence to support the adoption of AI into the UK breast screening programme. This thesis aims to fill the gaps in evidence highlighted in the NSC report for the performance of AI algorithms within a UK breast cancer screening population, as well as explore the various potential workflow deployment approaches of AI in the screening programme. I start by conducting a systematic review and meta-analysis of the current literature investigating the performance of stand-alone AI applications in breast cancer screening for detection and diagnosis as well as triage approaches. I then describe the creation of a large scale independent medical imaging database which is used in the studies throughout this thesis. The remainder of the thesis describes the results of three retrospective studies evaluating three different commercial AI algorithms. The first study assesses the ability of AI to detect interval cancers at the previous screen. The second study investigates the performance of AI as a stand-alone screen reader. The third study evaluates the proportion of cases identified for both high sensitivity rule out and high specificity rule in triage, as well as the proportion of cancers missed at these thresholds. Overall the results of this thesis will inform discussions around the use of AI in the UK breast screening programme as well as the design of future prospective trials.","abstract_html":"This thesis evaluates the application and performance of artificial intelligence (AI) in breast cancer screening. Breast cancer screening is conducted on a population scale using mammographic imaging for the earlier detection of breast cancer and has been shown to reduce mortality. A shortage of trained radiologists, as well as the demands of double reading, mean an approach to alleviate pressures within the breast screening workflow is sought. In addition, interval cancers occur at an estimated rate of 3.7/1000 women screened in the UK, thus methods to improve the sensitivity of screening and detect cancers earlier are also needed. Advances in AI over the past decade have demonstrated comparable performance to human readers and could provide a method for an adapted screening workflow to improve both efficiency and efficacy of screening. However, the 2021 National Screening Committee (NSC) report concluded that there was insufficient evidence to support the adoption of AI into the UK breast screening programme. This thesis aims to fill the gaps in evidence highlighted in the NSC report for the performance of AI algorithms within a UK breast cancer screening population, as well as explore the various potential workflow deployment approaches of AI in the screening programme. I start by conducting a systematic review and meta-analysis of the current literature investigating the performance of stand-alone AI applications in breast cancer screening for detection and diagnosis as well as triage approaches. I then describe the creation of a large scale independent medical imaging database which is used in the studies throughout this thesis. The remainder of the thesis describes the results of three retrospective studies evaluating three different commercial AI algorithms. The first study assesses the ability of AI to detect interval cancers at the previous screen. The second study investigates the performance of AI as a stand-alone screen reader. The third study evaluates the proportion of cases identified for both high sensitivity rule out and high specificity rule in triage, as well as the proportion of cancers missed at these thresholds. Overall the results of this thesis will inform discussions around the use of AI in the UK breast screening programme as well as the design of future prospective trials.","abstract_has_math":false,"creators":["Hickman, Sarah"],"institution":"University of Cambridge","degree_name":"Doctor of Philosophy (PhD)","degree_level":"Doctoral","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":["Gilbert, Fiona"],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-10-22","date_published":"2022-10-22","updated_at":"2026-07-22T22:24:14Z","subjects":["Artificial intelligence","Breast cancer screening","Mammography"],"languages":["eng"],"rights":[],"rights_urls":["https://www.rioxx.net/licenses/all-rights-reserved/"],"identifier_entries":[{"key":"dc:creator.authoridentifier","label":"Author Identifier","values":["0000000246377300"],"render_values":[{"text":"0000-0002-4637-7300","href":"https://orcid.org/0000-0002-4637-7300","code":true}]}]},"links":{"outbound_url":"https://doi.org/10.17863/CAM.89898","outbound_label":"DOI","outbound_source":"dc:identifier.doi"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Gilbert, Fiona"]},{"key":"dc:contributor.sponsor","label":"Sponsor","values":["Cancer Research UK"]},{"key":"dc:creator","label":"Author","values":["Hickman, Sarah"]},{"key":"dc:creator.authoridentifier","label":"Author Identifier","values":["0000000246377300"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.issued","label":"Date","values":["2022-10-22"]},{"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/1810/342480"]},{"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":["Artificial intelligence","Breast cancer screening","Mammography"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["https://www.rioxx.net/licenses/all-rights-reserved/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.doi","label":"DOI","values":["10.17863/CAM.89898"]},{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/2aea6bf9-d8d4-44ff-8794-07557e9685e7/download"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["This thesis evaluates the application and performance of artificial intelligence (AI) in breast cancer screening. Breast cancer screening is conducted on a population scale using mammographic imaging for the earlier detection of breast cancer and has been shown to reduce mortality. A shortage of trained radiologists, as well as the demands of double reading, mean an approach to alleviate pressures within the breast screening workflow is sought. In addition, interval cancers occur at an estimated rate of 3.7/1000 women screened in the UK, thus methods to improve the sensitivity of screening and detect cancers earlier are also needed. Advances in AI over the past decade have demonstrated comparable performance to human readers and could provide a method for an adapted screening workflow to improve both efficiency and efficacy of screening. However, the 2021 National Screening Committee (NSC) report concluded that there was insufficient evidence to support the adoption of AI into the UK breast screening programme. This thesis aims to fill the gaps in evidence highlighted in the NSC report for the performance of AI algorithms within a UK breast cancer screening population, as well as explore the various potential workflow deployment approaches of AI in the screening programme. I start by conducting a systematic review and meta-analysis of the current literature investigating the performance of stand-alone AI applications in breast cancer screening for detection and diagnosis as well as triage approaches. I then describe the creation of a large scale independent medical imaging database which is used in the studies throughout this thesis. The remainder of the thesis describes the results of three retrospective studies evaluating three different commercial AI algorithms. The first study assesses the ability of AI to detect interval cancers at the previous screen. The second study investigates the performance of AI as a stand-alone screen reader. The third study evaluates the proportion of cases identified for both high sensitivity rule out and high specificity rule in triage, as well as the proportion of cancers missed at these thresholds. Overall the results of this thesis will inform discussions around the use of AI in the UK breast screening programme as well as the design of future prospective trials."]},{"key":"dc:format.checksum.md5","label":"Dc Format Checksum Md5","values":["ac9facb3a4515d994da5652a8a751b36"]},{"key":"dc:title","label":"Title","values":["Evaluating Artificial Intelligence in Breast Cancer Screening"]}]}],"canonical_facts":{"dc:contributor.advisor":["Gilbert, Fiona"],"dc:contributor.sponsor":["Cancer Research UK"],"dc:creator":["Hickman, Sarah"],"dc:creator.authoridentifier":["0000000246377300"],"dc:date.issued":["2022-10-22"],"dc:description.abstract":["This thesis evaluates the application and performance of artificial intelligence (AI) in breast cancer screening. Breast cancer screening is conducted on a population scale using mammographic imaging for the earlier detection of breast cancer and has been shown to reduce mortality. A shortage of trained radiologists, as well as the demands of double reading, mean an approach to alleviate pressures within the breast screening workflow is sought. In addition, interval cancers occur at an estimated rate of 3.7/1000 women screened in the UK, thus methods to improve the sensitivity of screening and detect cancers earlier are also needed. Advances in AI over the past decade have demonstrated comparable performance to human readers and could provide a method for an adapted screening workflow to improve both efficiency and efficacy of screening. However, the 2021 National Screening Committee (NSC) report concluded that there was insufficient evidence to support the adoption of AI into the UK breast screening programme. This thesis aims to fill the gaps in evidence highlighted in the NSC report for the performance of AI algorithms within a UK breast cancer screening population, as well as explore the various potential workflow deployment approaches of AI in the screening programme. I start by conducting a systematic review and meta-analysis of the current literature investigating the performance of stand-alone AI applications in breast cancer screening for detection and diagnosis as well as triage approaches. I then describe the creation of a large scale independent medical imaging database which is used in the studies throughout this thesis. The remainder of the thesis describes the results of three retrospective studies evaluating three different commercial AI algorithms. The first study assesses the ability of AI to detect interval cancers at the previous screen. The second study investigates the performance of AI as a stand-alone screen reader. The third study evaluates the proportion of cases identified for both high sensitivity rule out and high specificity rule in triage, as well as the proportion of cancers missed at these thresholds. Overall the results of this thesis will inform discussions around the use of AI in the UK breast screening programme as well as the design of future prospective trials."],"dc:format.checksum.md5":["ac9facb3a4515d994da5652a8a751b36"],"dc:identifier.doi":["10.17863/CAM.89898"],"dc:identifier.uri":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/2aea6bf9-d8d4-44ff-8794-07557e9685e7/download"],"dc:language":["eng"],"dc:publisher.institution":["University of Cambridge"],"dc:relation.isreferencedby.uri":["https://www.repository.cam.ac.uk/handle/1810/342480"],"dc:rights":["https://www.rioxx.net/licenses/all-rights-reserved/"],"dc:subject":["Artificial intelligence","Breast cancer screening","Mammography"],"dc:title":["Evaluating Artificial Intelligence in Breast Cancer Screening"],"dc:type":["Thesis"],"dc:type.qualificationlevel":["Doctoral"],"dc:type.qualificationname":["Doctor of Philosophy (PhD)"]},"updated_at":"2026-07-22T22:24:14Z"}