Wen Li1,2, Yafen Li1,2, Wenjian Qin1, Xiaokun Liang1,2, Jianyang Xu3, Jing Xiong1, Yaoqin Xie1. 1. Institute of Biomedical and Health Engineering, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China. 2. Shenzhen Colleges of Advanced Technology, University of Chinese Academy of Sciences, Shenzhen 518055, China. 3. Shenzhen University General Hospital, Shenzhen University, Shenzhen 518055, China.
Abstract
BACKGROUND: Precise patient setup is critical in radiation therapy. Medical imaging plays an essential role in patient setup. As compared to computed tomography (CT) images, magnetic resonance image (MRI) has high contrast for soft tissues, which becomes a promising imaging modality during treatment. In this paper, we proposed a method to synthesize brain MRI images from corresponding planning CT (pCT) images. The synthetic MRI (sMRI) images can be used to align with positioning MRI (pMRI) equipped by an MRI-guided accelerator to account for the disadvantages of multi-modality image registration. METHODS: Several deep learning network models were applied to implement this brain MRI synthesis task, including CycleGAN, Pix2Pix model, and U-Net. We evaluated these methods using several metrics, including mean absolute error (MAE), mean squared error (MSE), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR). RESULTS: In our experiments, U-Net with L1+L2 loss achieved the best results with the lowest overall average MAE of 74.19 and MSE of 1.035*104, respectively, and produced the highest SSIM of 0.9440 and PSNR of 32.44. CONCLUSIONS: Quantitative comparisons suggest that the performance of U-Net, a supervised deep learning method, is better than the performance of CycleGAN, a typical unsupervised method, in our brain MRI synthesis procedure. The proposed method can convert pCT/pMRI multi-modality registration into mono-modality registration, which can be used to reduce registration error and achieve a more accurate patient setup. 2020 Quantitative Imaging in Medicine and Surgery. All rights reserved.
BACKGROUND: Precise patient setup is critical in radiation therapy. Medical imaging plays an essential role in patient setup. As compared to computed tomography (CT) images, magnetic resonance image (MRI) has high contrast for soft tissues, which becomes a promising imaging modality during treatment. In this paper, we proposed a method to synthesize brain MRI images from corresponding planning CT (pCT) images. The synthetic MRI (sMRI) images can be used to align with positioning MRI (pMRI) equipped by an MRI-guided accelerator to account for the disadvantages of multi-modality image registration. METHODS: Several deep learning network models were applied to implement this brain MRI synthesis task, including CycleGAN, Pix2Pix model, and U-Net. We evaluated these methods using several metrics, including mean absolute error (MAE), mean squared error (MSE), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR). RESULTS: In our experiments, U-Net with L1+L2 loss achieved the best results with the lowest overall average MAE of 74.19 and MSE of 1.035*104, respectively, and produced the highest SSIM of 0.9440 and PSNR of 32.44. CONCLUSIONS: Quantitative comparisons suggest that the performance of U-Net, a supervised deep learning method, is better than the performance of CycleGAN, a typical unsupervised method, in our brain MRI synthesis procedure. The proposed method can convert pCT/pMRI multi-modality registration into mono-modality registration, which can be used to reduce registration error and achieve a more accurate patient setup. 2020 Quantitative Imaging in Medicine and Surgery. All rights reserved.
Authors: Yannick Berker; Jochen Franke; André Salomon; Moritz Palmowski; Henk C W Donker; Yavuz Temur; Felix M Mottaghy; Christiane Kuhl; David Izquierdo-Garcia; Zahi A Fayad; Fabian Kiessling; Volkmar Schulz Journal: J Nucl Med Date: 2012-04-13 Impact factor: 10.057
Authors: Stefan Klein; Marius Staring; Keelin Murphy; Max A Viergever; Josien P W Pluim Journal: IEEE Trans Med Imaging Date: 2009-11-17 Impact factor: 10.048
Authors: Christoph Thilmann; Simeon Nill; Thomas Tücking; Angelika Höss; Bernd Hesse; Lars Dietrich; Rolf Bendl; Bernhard Rhein; Peter Häring; Christian Thieke; Uwe Oelfke; Juergen Debus; Peter Huber Journal: Radiat Oncol Date: 2006-05-24 Impact factor: 3.481
Authors: Reza Kalantar; Christina Messiou; Jessica M Winfield; Alexandra Renn; Arash Latifoltojar; Kate Downey; Aslam Sohaib; Susan Lalondrelle; Dow-Mu Koh; Matthew D Blackledge Journal: Front Oncol Date: 2021-07-30 Impact factor: 6.244