Sun Hongfei, Fan Rongbo, Li Chunying, Lu Zhengda, Xie Kai, Ni Xinye, Yang Jianhua
School of Automation, Northwestern Polytechnical University, Xi'an, China.
Department of Radiotherapy, Second People's Hospital of Changzhou, Nanjing Medical University, Changzhou, China.
Front Oncol. 2021 Mar 12;11:603844. doi: 10.3389/fonc.2021.603844. eCollection 2021.
To propose a synthesis method of pseudo-CT (CT) images based on an improved 3D cycle generative adversarial network (CycleGAN) to solve the limitations of cone-beam CT (CBCT), which cannot be directly applied to the correction of radiotherapy plans.
The improved U-Net with residual connection and attention gates was used as the generator, and the discriminator was a full convolutional neural network (FCN). The imaging quality of pseudo-CT images is improved by adding a 3D gradient loss function. Fivefold cross-validation was performed to validate our model. Each pseudo CT generated is compared against the real CT image (ground truth CT, CT) of the same patient based on mean absolute error (MAE) and structural similarity index (SSIM). The dice similarity coefficient (DSC) coefficient was used to evaluate the segmentation results of pseudo CT and real CT. 3D CycleGAN performance was compared to 2D CycleGAN based on normalized mutual information (NMI) and peak signal-to-noise ratio (PSNR) metrics between the pseudo-CT and CT images. The dosimetric accuracy of pseudo-CT images was evaluated by gamma analysis.
The MAE metric values between the CT and the real CT in fivefold cross-validation are 52.03 ± 4.26HU, 50.69 ± 5.25HU, 52.48 ± 4.42HU, 51.27 ± 4.56HU, and 51.65 ± 3.97HU, respectively, and the SSIM values are 0.87 ± 0.02, 0.86 ± 0.03, 0.85 ± 0.02, 0.85 ± 0.03, and 0.87 ± 0.03 respectively. The DSC values of the segmentation of bladder, cervix, rectum, and bone between CT and real CT images are 91.58 ± 0.45, 88.14 ± 1.26, 87.23 ± 2.01, and 92.59 ± 0.33, respectively. Compared with 2D CycleGAN, the 3D CycleGAN based pseudo-CT image is closer to the real image, with NMI values of 0.90 ± 0.01 and PSNR values of 30.70 ± 0.78. The gamma pass rate of the dose distribution between CT and CT is 97.0% (2%/2 mm).
The pseudo-CT images obtained based on the improved 3D CycleGAN have more accurate electronic density and anatomical structure.
提出一种基于改进的三维循环生成对抗网络（CycleGAN）的伪CT图像合成方法，以解决锥束CT（CBCT）不能直接应用于放射治疗计划校正的局限性。
将具有残差连接和注意力门的改进U-Net用作生成器，判别器为全卷积神经网络（FCN）。通过添加三维梯度损失函数提高伪CT图像的成像质量。进行五折交叉验证以验证我们的模型。基于平均绝对误差（MAE）和结构相似性指数（SSIM），将生成的每个伪CT与同一患者的真实CT图像（真实CT，CT）进行比较。使用骰子相似系数（DSC）系数评估伪CT和真实CT的分割结果。基于伪CT和CT图像之间的归一化互信息（NMI）和峰值信噪比（PSNR）指标，将三维CycleGAN性能与二维CycleGAN进行比较。通过伽马分析评估伪CT图像的剂量学准确性。
五折交叉验证中CT与真实CT之间的MAE指标值分别为52.03±4.26HU、50.69±5.25HU、52.48±4.42HU、51.27±4.56HU和51.65±3.97HU，SSIM值分别为0.87±0.02、0.86±0.03、0.85±0.02、0.85±0.03和0.87±0.03。CT与真实CT图像之间膀胱、宫颈、直肠和骨骼分割的DSC值分别为91.58±0.45、88.14±1.26、87.23±2.01和92.59±0.33。与二维CycleGAN相比，基于三维CycleGAN的伪CT图像更接近真实图像，NMI值为0.90±0.01，PSNR值为30.70±0.78。CT与CT之间剂量分布的伽马通过率为97.0%（2%/2mm）。
基于改进的三维CycleGAN获得的伪CT图像具有更准确的电子密度和解剖结构。
