ZusammenfassungWith the growing number of studies in clinical, scientific and industrial settings, the identification and accurate assessment of incidental findings in MRI imaging is of increasing importance. Deep-learning and density-estimation based methods have successfully been used for anomaly detection and localization in smaller medical image datasets. The aim of this study is to explore the potential of these highly promising techniques for the automatic identification and evaluation of incidental findings in large-scale population studies like NAKO

Schlüsselwörter-

EinrichtungenDKFZ, Universitätsklinikum Freiburg, Universität Tübingen, DKFZ Heidelberg, University Hospital Heidelberg, Deutsches Krebsforschungszentrum (DKFZ), Diagnostische und Interventionelle Radiologie, Universitätsklinik Heidelberg, Radiologische Klinik, Klinik für Diagnostische und Interventionelle Radiologie, Forschungszentrum Jülich