ZusammenfassungChronic obstructive pulmonary disease (COPD) is one of the most important non-communicable disease entities among respiratory diseases, recognized by the World Health Organization and the European Union and is a significant cause of morbidity and mortality. However, in respiratory health assessment based on large population-based studies, wide variations in mortality and admission rates for respiratory conditions have been observed across Western and even European countries. Beside pulmonary function test, (PFT) imaging plays an increasing role in evaluating pulmonary disease. Although CT is the mainstay of lung imaging, magnetic resonance imaging (MRI) has become a reliable, radiation-free alternative in clinical settings and research, increasingly used in population-based cohort studies. Even though whole-body MR is often not dedicated for lung imaging, the lung is part of the covered region of interest. Within the NAKO, we aim to develop an automated segmentation algorithm included into an analysis pipeline including dedicated quality assurance. This will derive lung boundaries and subsequently lung volumes, separated for the left and right. At a later stage, this will allow us to extent our knowledge regarding the association of lung volumes derived by non-dedicated whole-body MRI with risk factors, pulmonary function test and COPD. In addition, we aim to use Deep Learning techniques together with Radiomic Feature (e.g. of the pulmonary vascular bed) for better prediction models for COPD based on non-dedicated whole-body MRI. Future results may suggest, that non-dedicated whole-body MRI could be an accurate, radiation-free, and possibly one-stop-shop imaging tool, used in novel radiomics and personalized medicine with potential for assessment of pulmonary health.

Schlüsselwörter-

EinrichtungenUniversitätsklinikum Freiburg, Helmholtz Zentrum München, Universitätsklinikum Heidelberg, University Hospital Heidelberg, DKFZ, Uniklinikum Heidelberg, Diagnostische und Interventionelle Radiologie, Radiologie/Uni Heidelberg, Klinikum der Universität München, Klinik für Diagnostische und interventionelle Radiologie Freiburg, DKFZ Heidelberg