• Login
    View Item 
    •   Mak UD Home
    • College of Computing and Information Sciences (CoCIS)
    • School of Computing and Informatics Technology (CIT)
    • School of Computing and Informatics Technology Collection
    • View Item
    •   Mak UD Home
    • College of Computing and Information Sciences (CoCIS)
    • School of Computing and Informatics Technology (CIT)
    • School of Computing and Informatics Technology Collection
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    RaiD - A system for automated abnormality detection in chest radiographs

    Thumbnail
    View/Open
    Undergraduate dissertation (5.205Mb)
    Date
    2022-01
    Author
    Mugisha, Stephen
    Mayanja, Benjamin Vincent
    Katwere, Leo
    Kengo, Wada
    Metadata
    Show full item record
    Abstract
    Medical imaging techniques such as X-rays, Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) scans are widely used to diagnose several medical conditions ranging from broken bones to detecting cancer tumors in the body. Traditionally, radiologists take radiographs from imaging machines and manually (using their naked eye) analyze them to detect any abnormalities based on their knowledge of what a normal radiograph without any defects looks like. This process is of course error-prone even for the most experienced radiologists especially in the case of chest radiographs. Some existing machine learning techniques categorize the images into lists of findings without proper localization of abnormalities in an input image/radiograph thus creating another challenge in the interpretation of results for radiologists. These methods have also been shown to be less accurate than their deep learning counterparts. RaiD, automates the process of abnormality detection in chest radiographs for specific classes of abnormalities through the use of a powerful deep learning algorithm called Yolo. Deep learning techniques have proven to be more accurate than traditional machine learning methods on a variety of tasks and thus justifying our need to apply them. This improved efficiency will consequently lead to a reduction in errors in diagnosis.
    URI
    http://hdl.handle.net/20.500.12281/11710
    Collections
    • School of Computing and Informatics Technology Collection

    DSpace 5.8 copyright © Makerere University 
    Contact Us | Send Feedback
    Theme by 
    Atmire NV
     

     

    Browse

    All of Mak UDCommunities & CollectionsTitlesAuthorsBy AdvisorBy Issue DateSubjectsBy TypeThis CollectionTitlesAuthorsBy AdvisorBy Issue DateSubjectsBy Type

    My Account

    LoginRegister

    Statistics

    Most Popular ItemsStatistics by CountryMost Popular Authors

    DSpace 5.8 copyright © Makerere University 
    Contact Us | Send Feedback
    Theme by 
    Atmire NV