Department of Biomedical Imaging and Radiological Sciences, National Yang- Ming University, B306, Experimental building No. 155, Sec. 2, Linong Street, Taipei, 112, Taiwan.
Division of Radiation Oncology, Department of Radiology, Far Eastern Memorial Hospital, New Taipei City, Taiwan.
Sci Rep. 2021 Jan 13;11(1):1133. doi: 10.1038/s41598-020-80803-2.
Cone-beam computed tomography (CBCT) integrated with a linear accelerator is widely used to increase the accuracy of radiotherapy and plays an important role in image-guided radiotherapy (IGRT). For comparison with fan-beam computed tomography (FBCT), the image quality of CBCT is indistinct due to X-ray scattering, noise, and artefacts. We proposed a deep learning model, "Cycle-Deblur GAN", combined with CycleGAN and Deblur-GAN models to improve the image quality of chest CBCT images. The 8706 CBCT and FBCT image pairs were used for training, and 1150 image pairs were used for testing in deep learning. The generated CBCT images from the Cycle-Deblur GAN model demonstrated closer CT values to FBCT in the lung, breast, mediastinum, and sternum compared to the CycleGAN and RED-CNN models. The quantitative evaluations of MAE, PSNR, and SSIM for CBCT generated from the Cycle-Deblur GAN model demonstrated better results than the CycleGAN and RED-CNN models. The Cycle-Deblur GAN model improved image quality and CT-value accuracy and preserved structural details for chest CBCT images.
锥形束计算机断层扫描（CBCT）与线性加速器相结合被广泛用于提高放射治疗的准确性，在图像引导放射治疗（IGRT）中发挥着重要作用。与扇形束计算机断层扫描（FBCT）相比，由于 X 射线散射、噪声和伪影，CBCT 的图像质量较差。我们提出了一种深度学习模型“Cycle-Deblur GAN”，结合了 CycleGAN 和 Deblur-GAN 模型，以提高胸部 CBCT 图像的质量。使用 8706 对 CBCT 和 FBCT 图像对进行训练，在深度学习中使用 1150 对图像对进行测试。与 CycleGAN 和 RED-CNN 模型相比，Cycle-Deblur GAN 模型生成的 CBCT 图像在肺、乳房、纵隔和胸骨处的 CT 值更接近 FBCT。Cycle-Deblur GAN 模型生成的 CBCT 的 MAE、PSNR 和 SSIM 的定量评估结果优于 CycleGAN 和 RED-CNN 模型。Cycle-Deblur GAN 模型提高了胸部 CBCT 图像的质量和 CT 值准确性，并保留了结构细节。
