[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123334-en":3,"doc-seo-123334-105":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},123334,1374391974468,"Eden","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Mass Detection In Mammograms Using Computer Vision And Machine Learning - Thesis Overview","This study explores mammogram mass segmentation using computer vision and machine learning techniques while avoiding neural networks. The research addresses limitations of neural approaches that require substantial training data and computing resources, which can hinder real-time CAD deployment. The proposed pipeline includes Otsu-threshold binarization for background noise removal, contrast improvement via histogram equalization, segmentation with Gaussian Mixture Models, and ROI-based abnormality extraction with false-positive reduction through classification algorithms. Sensitivity is prioritized to match application needs, aiming for resource-efficient alternatives when neural networks are impractical.","Mass Detection In Mammograms Using Computer Vision And Machine  \nLearning  \nA Thesis submitted to the faculty of  \nSan Francisco State University  \nIn partial satisfaction of the  \nrequirements for  \nthe Degree  \nMasters of Science  \nin  \nStatistical Data Science  \nby  \nSindhu Muralidhara Havaldar  \nSan Francisco, California  \nDecember 2023  \nCopyright by  \nSindhu Muralidhara Havaldar  \n2023  \nCertification of Approval  \nI certify that I have read Mass Detection In Mammograms Using Computer Vision And Machine Learning by Sindhu Muralidhara Havaldar and that in my opinion this work meets the criteria for approving a thesis submitted in partial fulfillment of the requirement for the degree Masters of Science at San Francisco State University.  \nAlexandra Piryatinska, Ph.D Professor  \nThesis Committee Chair  \nMohammad Kafai, MS Professor  \nAbstract  \nThis study explores mammogram mass segmentation, focusing on computer vision and machine learning methods while avoiding neural networks. Despite their exceptional performance, neural networks demand substantial resources and data, presenting challenges. Our research aims to uncover the potential of non-neural models for effective mass segmentation in mammograms. This approach is not resource intensive and can be used for real-time mass segmentation in CAD systems. The steps involved in this process include: eliminating background noise through binarization through otsu-thresholding, enhancing images using a unique histogram equalization method,segmenting mammograms with Gaussian Mixture models to isolate crucial patterns, extracting potential abnormalities and filtering falsepositives using classification algorithms.  \nNotably, we prioritize sensitivity over specificity, tailored to our application’s needs. This thorough exploration into non-neural models intends to provide insights into their effectiveness for mammogram mass segmentation, offering potential resource-efficient alternatives in situations where neural networks are impractical.  \nv  \nAcknowledgments  \nI extend my heartfelt gratitude to Professor Dr. Alexandra Piryatinska for her consistent guidance in unraveling the complex concepts presented in research papers. Her encouragement to delve into my interests and her unwavering support during the thesis process have been immensely valuable. Additionally, I would like to convey my thanks to Professor Dr. Mohammad Kafai for the thorough review of my work and the insightful feedback he provided. The combined mentorship of these professors has greatly enhanced my academic experience, and I appreciate their contributions to the successful culmination of this project.  \nvi  \nTable of Contents  \nTable of Contents vi  \nList of Tables vii  \nList of Figures viii  \n1 Overview of Segmentation Methods 4  \n1.1 Background Noise Removal ........................... 4  \n1.2 Pectoral Muscle Removal ............................. 6  \n1.3 Contrast Enhancement .............................. 7  \n1.4 Segmentation ................................... 12  \n1.5 Edge Detection .................................. 14  \n1.6 ROI Extraction & Filtering ........................... 17  \n2 Methodology 21  \n2.1 Data Description ................................. 21  \n2.2 Data Preprocessing ................................ 23  \n2.3 Segmentation ................................... 24  \n2.4 ROI Extraction .................................. 27  \n2.5 ROI Feature Extraction & Classification .................... 27  \n3 Results & Conclusion 30  \n3.1 Results ....................................... 30  \n3.2 Conclusion & Future Work ............................ 32  \nBibliography 35  \nAppendices 38  \n.1 Python Code Block ................................ 38  \nvii  \nList of Tables  \n2.1 Train and Test Data Sizes for Different Tissue Types ............... 22  \n3.1 Breast Tissue Segmentation and Detection Results ................ 31  \n3.2 Performance Metrics for Four Models ........................ 32  \nviii  \nList of Figures  ","cbCait90fTlQM5Aw","https://ap.wps.com/l/cbCait90fTlQM5Aw","pdf",5121696,1,90,"English","en",105,"# Overview of Segmentation Methods\n## Background Noise Removal\n## Pectoral Muscle Removal\n## Contrast Enhancement\n## Segmentation\n## Edge Detection\n## ROI Extraction & Filtering\n# Methodology\n## Data Description\n## Data Preprocessing\n## Segmentation\n## ROI Extraction\n## ROI Feature Extraction & Classification\n# Results & Conclusion\n## Results\n## Conclusion & Future Work","[{\"question\":\"Why does the thesis avoid neural networks for mammogram mass segmentation?\",\"answer\":\"Neural network methods require large amounts of data and computing resources, making real-time use difficult. The thesis investigates whether non-neural models can achieve effective segmentation with lower resource demands.\"},{\"question\":\"What key steps are included in the proposed segmentation pipeline?\",\"answer\":\"The pipeline performs background noise removal with Otsu-threshold binarization, enhances images using histogram equalization, segments mammograms with Gaussian Mixture Models, and then extracts abnormalities and filters false positives using classification algorithms.\"},{\"question\":\"How does the thesis evaluate performance and what metric preference is used?\",\"answer\":\"The approach prioritizes sensitivity over specificity, aligning with the needs of the intended application. Results include performance metrics comparing different models.\"}]","Mass Detection In Mammograms Using Computer Vision And Machine Learning - Thesis Overview | PDF",1785815989,227,{"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},"mass-detection-in-mammograms-using-computer-vision-and-machine-learning-thesis-overview","",{"@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/mass-detection-in-mammograms-using-computer-vision-and-machine-learning-thesis-overview/123334/",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","2026-08-04",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},"Why does the thesis avoid neural networks for mammogram mass segmentation?","Question",{"text":75,"@type":76},"Neural network methods require large amounts of data and computing resources, making real-time use difficult. The thesis investigates whether non-neural models can achieve effective segmentation with lower resource demands.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What key steps are included in the proposed segmentation pipeline?",{"text":80,"@type":76},"The pipeline performs background noise removal with Otsu-threshold binarization, enhances images using histogram equalization, segments mammograms with Gaussian Mixture Models, and then extracts abnormalities and filters false positives using classification algorithms.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the thesis evaluate performance and what metric preference is used?",{"text":84,"@type":76},"The approach prioritizes sensitivity over specificity, aligning with the needs of the intended application. Results include performance metrics comparing different models.","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},[93,96,100,104,109,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":21,"slug":95},"Story & Novel","story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":46,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]