Tom Russ1, Stephan Goerttler2, Alena-Kathrin Schnurr2, Dominik F Bauer2, Sepideh Hatamikia3,4, Lothar R Schad2, Frank G Zöllner2, Khanlian Chung2. 1. Computer Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany. tom.russ@medma.uni-heidelberg.de. 2. Computer Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany. 3. Austrian Center for Medical Innovation and Technology, Vienna, Austria. 4. Center for Medical Physics and Biomedical Engineering, Medical University Vienna, Vienna, Austria.
Abstract
PURPOSE: The potential of medical image analysis with neural networks is limited by the restricted availability of extensive data sets. The incorporation of synthetic training data is one approach to bypass this shortcoming, as synthetic data offer accurate annotations and unlimited data size. METHODS: We evaluated eleven CycleGAN for the synthesis of computed tomography (CT) images based on XCAT body phantoms. The image quality was assessed in terms of anatomical accuracy and realistic noise properties. We performed two studies exploring various network and training configurations as well as a task-based adaption of the corresponding loss function. RESULTS: The CycleGAN using the Res-Net architecture and three XCAT input slices achieved the best overall performance in the configuration study. In the task-based study, the anatomical accuracy of the generated synthetic CTs remained high ([Formula: see text] and [Formula: see text]). At the same time, the generated noise texture was close to real data with a noise power spectrum correlation coefficient of [Formula: see text]. Simultaneously, we observed an improvement in annotation accuracy of 65% when using the dedicated loss function. The feasibility of a combined training on both real and synthetic data was demonstrated in a blood vessel segmentation task (dice similarity coefficient [Formula: see text]). CONCLUSION: CT synthesis using CycleGAN is a feasible approach to generate realistic images from simulated XCAT phantoms. Synthetic CTs generated with a task-based loss function can be used in addition to real data to improve the performance of segmentation networks.
PURPOSE: The potential of medical image analysis with neural networks is limited by the restricted availability of extensive data sets. The incorporation of synthetic training data is one approach to bypass this shortcoming, as synthetic data offer accurate annotations and unlimited data size. METHODS: We evaluated eleven CycleGAN for the synthesis of computed tomography (CT) images based on XCAT body phantoms. The image quality was assessed in terms of anatomical accuracy and realistic noise properties. We performed two studies exploring various network and training configurations as well as a task-based adaption of the corresponding loss function. RESULTS: The CycleGAN using the Res-Net architecture and three XCAT input slices achieved the best overall performance in the configuration study. In the task-based study, the anatomical accuracy of the generated synthetic CTs remained high ([Formula: see text] and [Formula: see text]). At the same time, the generated noise texture was close to real data with a noise power spectrum correlation coefficient of [Formula: see text]. Simultaneously, we observed an improvement in annotation accuracy of 65% when using the dedicated loss function. The feasibility of a combined training on both real and synthetic data was demonstrated in a blood vessel segmentation task (dice similarity coefficient [Formula: see text]). CONCLUSION: CT synthesis using CycleGAN is a feasible approach to generate realistic images from simulated XCAT phantoms. Synthetic CTs generated with a task-based loss function can be used in addition to real data to improve the performance of segmentation networks.
Authors: Geert Litjens; Thijs Kooi; Babak Ehteshami Bejnordi; Arnaud Arindra Adiyoso Setio; Francesco Ciompi; Mohsen Ghafoorian; Jeroen A W M van der Laak; Bram van Ginneken; Clara I Sánchez Journal: Med Image Anal Date: 2017-07-26 Impact factor: 8.545
Authors: Camilo Bermudez; Andrew J Plassard; Taylor L Davis; Allen T Newton; Susan M Resnick; Bennett A Landman Journal: Proc SPIE Int Soc Opt Eng Date: 2018-03
Authors: Pedro Costa; Adrian Galdran; Maria Ines Meyer; Meindert Niemeijer; Michael Abramoff; Ana Maria Mendonca; Aurelio Campilho Journal: IEEE Trans Med Imaging Date: 2017-10-02 Impact factor: 10.048
Authors: Frank G Zöllner; Amira Šerifović-Trbalić; Gordian Kabelitz; Marek Kociński; Andrzej Materka; Peter Rogelj Journal: MAGMA Date: 2019-10-09 Impact factor: 2.310
Authors: Ehsan Abadi; William P Segars; Benjamin M W Tsui; Paul E Kinahan; Nick Bottenus; Alejandro F Frangi; Andrew Maidment; Joseph Lo; Ehsan Samei Journal: J Med Imaging (Bellingham) Date: 2020-04-11
Authors: Dominik F Bauer; Tom Russ; Barbara I Waldkirch; Christian Tönnes; William P Segars; Lothar R Schad; Frank G Zöllner; Alena-Kathrin Golla Journal: Int J Comput Assist Radiol Surg Date: 2021-05-02 Impact factor: 3.421