Radiol Artif Intell 2020 May 27;2(3):e190035. Epub 2020 May 27.
Service of Radiology, Department of Diagnostics, Geneva University Hospital, Rue Gabrielle Perret-Gentil 4, 1211 Geneva 14, Switzerland (S.M., J.H., S.P.M., X.M.); Department of Radiology and Medical Informatics, University of Geneva, Geneva, Switzerland (S.M., J.H., M.G.P., S.P.M., D.V.D.V., X.M.); and Institute of Bioengineering/Center for Neuroprosthetics, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland (S.M., M.G.P., D.V.D.V.).
Purpose: To assess the contribution of a generative adversarial network (GAN) to improve intermanufacturer reproducibility of radiomic features (RFs).
Materials And Methods: The authors retrospectively developed a cycle-GAN to translate texture information from chest radiographs acquired using one manufacturer (Siemens) to chest radiographs acquired using another (Philips), producing fake chest radiographs with different textures. The authors prospectively evaluated the ability of this texture-translation cycle-GAN to reduce the intermanufacturer variability of RFs extracted from the lung parenchyma. Read More