Upgrade to Pro — share decks privately, control downloads, hide ads and more …

MIWAD: A software suite for building CAD methods

CETA-Ciemat
July 27, 2012
140

MIWAD: A software suite for building CAD methods

Slides from the Symposium "Methods in Breast Cancer Aided Diagnosis" Porto (Portugal)

CETA-Ciemat

July 27, 2012
Tweet

Transcript

  1. MIWAD: A software suite
    for building CAD methods
    Speaker: José M. Franco Valiente
    “METHODS IN BREAST CANCER COMPUTER AIDED DIAGNOSIS”
    ICEM 15 Porto 27-07-2012

    View full-size slide

  2. Authors
    CETA-CIEMAT: José M. Franco Valiente, César
    Suárez Ortega, Manuel Rubio del Solar,
    Guillermo Díaz Herrero, Raúl Ramos Pollán
    INEGI-FEUP: Miguel A. Guevara López, Naimy
    González de Posada, Daniel C. Moura, Pedro
    Cunha
    FMUP-HSJ: Isabel M. A. Pereira Ramos, Joana
    Pinheiro Loureiro, Teresa Cardoso
    Fernandes, Bruno M. Ferreira de Araújo

    View full-size slide

  3. The Project
    Imed
    Exploiting e-Infrastructures
    for research in breast
    cancer CAD methods

    View full-size slide

  4. Current Results
    BCDR: Breast Cancer Digital Repository
    MIWAD: Mammography Image
    Workstation for Analysis and Diagnosis
    MLC: A first set of Machine Learning
    Classifiers

    View full-size slide

  5. MIWAD
    A novel software suite for
    processing, analyzing and
    diagnosing mammography
    Images
    MIWAD DB + MIWAD CAD

    View full-size slide

  6. MIWAD Storage Layer: DRI

    View full-size slide

  7. MIWAD DB
    Manages BCDR content
    Specialists can check clinical data
    Used as a gateway to store training data
    Executes MLC classifiers

    View full-size slide

  8. MIWAD CAD
    Implements image processing algorithms
    Creates segmentations to classify
    Calculates features
    Based on TUDOR Viewer

    View full-size slide

  9. MIWAD Execution Modes
    Training mode: Used to gather
    information for the training of the MLCs.
    Diagnosis mode: Provide users a second
    opinion for new cases.

    View full-size slide

  10. MIWAD Lesion Workflow

    View full-size slide

  11. Classifiers integration
    MIWAD Classifiers API

    View full-size slide

  12. Future work
    Including new pre-processing operations
    Other image modalities support
    Enhancing browsing functionalities
    DICOM importation / exportation
    Improvement of the Classifiers API

    View full-size slide

  13. Future work
    MIWAD development is supervised and
    validated by radiologists from
    the FMUP-HSJ so the evolution of the
    software is highly oriented to fit their
    necessities

    View full-size slide

  14. Demo
    No demo, sorry 

    View full-size slide

  15. Demo
    No demo, sorry 
    But it is REAL!

    View full-size slide

  16. Conclusion
    MIWAD is used by radiologists to
    evaluate and classify the cases from BCDR
    MIWAD is used by researchers to build
    and evaluate MLCs
    Third party classifiers can be easily
    integrated as a plugin

    View full-size slide

  17. Acknowledgements
    The three institutions express their
    gratitude for the support of the European
    Regional Development Fund (FEDER-
    ERDF)

    View full-size slide

  18. The End
    Thanks for your attention!

    View full-size slide

  19. The End
    Thanks for your attention!
    Any Question?

    View full-size slide