OPTIMAM Image Database: Version 2

  • Research type

    Research Database

  • IRAS ID

    265403

  • Contact name

    Kenneth Young

  • Contact email

    ken.young@nhs.net

  • Research summary

    OPTIMAM database of digital mammograms for the optimisation of breast cancer detection

  • REC name

    South Central - Oxford C Research Ethics Committee

  • REC reference

    19/SC/0284

  • Date of REC Opinion

    12 Jul 2019

  • REC opinion

    Favourable Opinion

  • Data collection arrangements

    The data comprises de-identified digital mammograms and de-identified associated radiological and pathological data. Images and data are generally collected by an automated system which identifies screen detected cancers prospectively at the main clinical sites. Some digital mammograms with normal and benign pathology are also included in the database.

  • Research programme

    The data is used for research and training. The initial purpose for collecting the data was to conduct research described in the OPTIMAM2 Research Project funded by Cancer Research UK. This research evaluated how various technical factors affect cancer detection using mammographic images. This was achieved by experiments on the ability of radiologists to detect cancers in studies which closely mimic breast screening by requiring search and detection in either real or virtual clinical images, i.e. real patient backgrounds or projections of anthropomorphic phantoms enriched with simulated or real lesions. The research provides information on how best to use the latest digital X-ray imaging technology. The de-identified images and data are also made available to other researchers using a written agreement limiting the use of the data to specific projects. Much recent research using the database by our own team and others helps develop and evaluate artificial intelligence tools to automate the detection of breast cancers. This is important because even small improvements in cancer detection in breast screening can lead to large numbers of extra cancers being detected and fewer unnecessary recalls as well as more efficient use of scarce resources e.g. radiologists’ time.

  • Research database title

    OPTIMAM database of digital mammograms for the optimisation of breast cancer detection

  • Establishment organisation

    Royal Surrey County Hospital

  • Establishment organisation address

    Egerton Road

    Department of Medical Physics

    Guildford

    GU2 7XX