[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-125237-en":3,"doc-seo-125237-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},125237,1099514067438,"River Wang","https://ap-avatar.wpscdn.com/avatar/100002539ee87300030?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780474512215547542",8,"Research & Report","Beyond the Algorithm - Evaluating Machine Learning for Breast Cancer Screening","The study evaluates machine learning (ML) for breast cancer screening using a dual method: radiologist interviews and empirical model testing. Persistent racial disparities in outcomes motivate analysis of real-world AI adoption, where clinician perspectives reflect prior technological shifts and reveal major economic, legal, and regulatory barriers. Model assessment of the GMIC approach on the racially diverse EMBED dataset shows uneven performance: better malignant classification for White patients and better benign classification for Black patients, suggesting the algorithm prioritizes differently across groups. Findings support fairness-aware models, diverse training data, and clearer regulatory frameworks, enabling responsible, interdisciplinary implementation with proper reimbursement and deployment guidelines.","BEYOND THE ALGORITHM: EVALUATING MACHINE LEARNING FOR BREAST  \nCANCER SCREENING  \nMia E. Martinez  \nTC 660H  \nPlan II Honors Program  \nThe University of Texas at Austin  \nApril 25, 2025  \nEdward Castillo, Ph.D. Department of Biomedical Engineering Supervising Professor  \nMia Markey, Ph.D. Department of Biomedical Engineering  \nSecondary Reader  \nABSTRACT  \nName: Mia E. Martinez  \nTitle: Beyond the Algorithm: Evaluating Machine Learning for Breast Cancer Screening  \nSupervising Professor: Edward Castillo, Ph.D.  \nThis study explores machine learning (ML) applications in breast cancer screening through a dual approach of radiologist interviews and empirical model evaluation. Despite improved survival rates, breast cancer outcomes show persistent racial disparities, particularly between Black and White patients. Our thematic analysis of radiologist interviews revealed key insights into the clinical adoption of AI systems, highlighting how perspectives are shaped by previous technological transitions and identifying significant economic, legal, and regulatory barriers to implementation. Meanwhile, evaluation of the GMIC model on the racially diverse EMBED dataset demonstrated concerning performance disparities-the model had better malignant classification for White patients, but better benign classification for Black patients. These results illustrate a clinically significant situation where the algorithm prioritizes different things for Black and White patients. These findings underscore the need for more diverse training datasets, fairness-aware algorithms, and clearer regulatory frameworks to ensure that ML-enhanced breast cancer screening reduces rather than reinforces existing healthcare disparities. Successful implementation will require interdisciplinary collaboration between clinicians, developers, healthcare systems, and policymakers to overcome skepticism, establish appropriate reimbursement structures, and develop guidelines for responsible deployment of AI in clinical settings.  \nACKNOWLEDGEMENTS  \nFirst, I would like to express my gratitude for the support of my two thesis advisors, Dr. Castillo and Dr. Markey. As experts in the field, you have provided invaluable knowledge, insight, and resources. Each of you demonstrated an unwavering level of support for me. Moreover, you have allowed me to grow as a learner and researcher. I am deeply grateful foryour mentorship.  \nAdditionally, I would like to thank the graduate students who guided me through my analytical processes, specifically Alaa Melek and Tracey Liu. Thank you for taking the time to act as a source of guidance and support over the past academic year. This project could not have been completed without you.  \nTo my mom and dad, I would not be where I am without you guys. You have taught me the importance of an education, and I will never take for granted the opportunities you have worked so hard to provide for me. As a young girl, you cultivated my love for science and math and ensured I knew I could do anything I set my mind to. Thank you for believing in me.  \nTable of Contents  \nABSTRACT..........................................................................................................................................2  \nACKNOWLEDGEMENTS ................................................................................................................3  \nINTRODUCTION................................................................................................................................ 5  \nBACKGROUND .................................................................................................................................. 7  \nSCREENING PROCEDURE: MAMMOGRAMS ...........................................................................................7  \nSCREENING PROCEDURE: RISK ASSESSMENT.......................................................................................9  \nDISPARITIES IN DIAGNOSIS AND OUTCOMES ............................","cbCaivizaM2XTMsH","https://ap.wps.com/l/cbCaivizaM2XTMsH","pdf",648201,1,53,"English","en",105,"# Abstract\n# Acknowledgements\n# Introduction\n# Background\n## Screening Procedure: Mammograms\n## Screening Procedure: Risk Assessment\n# Disparities in Diagnosis and Outcomes\n# Potential for Artificial Intelligence and Machine Learning\n# Literature Review\n## Objective 1: Improving Accuracy, Sensitivity, and Specificity of Detection Models\n## Objective 2: Employing Hybrid DL Models for Risk Prediction\n## Objective 3: Improving the Ability of Models to Generalize on a Diverse Population\n## Objective 4: Translating Success into Clinical Breast Cancer Screening Programs\n# Research Methods\n## Interviews: Subject Selection\n## Interviews: Data Collection\n## Interviews: Data Analysis\n## Performance Assessment: Model and Dataset Selection\n# Analytical Chapter 1: Interviews with Radiologists\n## Technological Evolution of Mammogram Interpretation\n## Barriers to AI Implementation in Clinical Practice","[{\"question\":\"How does the study evaluate machine learning for breast cancer screening?\",\"answer\":\"It combines radiologist interviews with empirical evaluation of an ML model. The interview analysis examines clinical adoption of AI systems, while the model evaluation tests performance on a racially diverse dataset.\"},{\"question\":\"What disparity-related findings emerge from radiologist interview themes?\",\"answer\":\"The thematic analysis highlights adoption perspectives shaped by earlier technological transitions and identifies economic, legal, and regulatory barriers. It also connects these factors to the broader challenge of persistent racial disparities in outcomes.\"},{\"question\":\"What performance differences are observed when evaluating the GMIC model?\",\"answer\":\"On the EMBED dataset, the model performs differently across racial groups: it shows better malignant classification for White patients and better benign classification for Black patients. This indicates uneven prioritization across groups.\"}]","Beyond the Algorithm - Evaluating Machine Learning for Breast Cancer Screening | PDF",1785897656,134,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"beyond-the-algorithm-evaluating-machine-learning-for-breast-cancer-screening","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/beyond-the-algorithm-evaluating-machine-learning-for-breast-cancer-screening/125237/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"How does the study evaluate machine learning for breast cancer screening?","Question",{"text":75,"@type":76},"It combines radiologist interviews with empirical evaluation of an ML model. The interview analysis examines clinical adoption of AI systems, while the model evaluation tests performance on a racially diverse dataset.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What disparity-related findings emerge from radiologist interview themes?",{"text":80,"@type":76},"The thematic analysis highlights adoption perspectives shaped by earlier technological transitions and identifies economic, legal, and regulatory barriers. It also connects these factors to the broader challenge of persistent racial disparities in outcomes.",{"name":82,"@type":73,"acceptedAnswer":83},"What performance differences are observed when evaluating the GMIC model?",{"text":84,"@type":76},"On the EMBED dataset, the model performs differently across racial groups: it shows better malignant classification for White patients and better benign classification for Black patients. This indicates uneven prioritization across groups.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]