Deep metric learning for multi-labelled radiographs

Mauro Annarumma, Giovanni Montana

    Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review

    8 Citations (Scopus)

    Abstract

    Many radiological studies can reveal the presence of several co-existing abnormalities, each one represented by a distinct visual pattern. In this article we address the problem of learning a distance metric for plain radiographs that captures a notion of "radiological similarity": two chest radiographs are considered to be similar if they share similar abnormalities. Deep convolutional neural networks (DCNs) are used to learn a low-dimensional embedding for the radiographs that is equipped with the desired metric. Two loss functions are proposed to deal with multi-labelled images and potentially noisy labels. We report on a large-scale study involving over 745,000 chest radiographs whose labels were automatically extracted from free-text radiological reports through a natural language processing system. Using 4,500 validated exams, we demonstrate that the methodology performs satisfactorily on clustering and image retrieval tasks.

    Original languageEnglish
    Title of host publicationProceedings of the 33rd Annual ACM Symposium on Applied Computing, SAC 2018
    PublisherAssociation for Computing Machinery
    Pages34-37
    Number of pages4
    ISBN (Electronic)9781450351911
    DOIs
    Publication statusPublished - 9 Apr 2018
    Event33rd Annual ACM Symposium on Applied Computing, SAC 2018 - Pau, France
    Duration: 9 Apr 201813 Apr 2018

    Publication series

    NameProceedings of the ACM Symposium on Applied Computing

    Conference

    Conference33rd Annual ACM Symposium on Applied Computing, SAC 2018
    Country/TerritoryFrance
    CityPau
    Period9/04/201813/04/2018

    Keywords

    • Convolutional networks
    • Deep metric learning
    • X-rays

    Fingerprint

    Dive into the research topics of 'Deep metric learning for multi-labelled radiographs'. Together they form a unique fingerprint.

    Cite this