We propose a learning method to generate synthetic CT (sCT) image for MRI-only radiation treatment planning. Theproposed method integrated a dense-block concept into a cycle-generative adversarial network (cycle-GAN) framework,which is named as dense-cycle-GAN in this study. Compared with GAN, the cycle-GAN includes an inversetransformation between CT (ground truth) and sCT, which could further constrain the learning model. A 2.5D fullyconvolution neural network (FCN) with dense-block was introduced in generator to enable end-to-end transformation. AFCN is used in discriminator to urge the generator’s sCT to be similar with the ground-truth CT images. The well-trainedmodel was used to generate the sCT of a new MRI. This proposed algorithm was evaluated using 14 patients’ data withboth MRI and CT images. The mean absolute error (MAE), peak signal-to-noise ratio (PSNR) and normalized crosscorrelation (NCC) indexes were used to quantify the correction accuracy of the prediction algorithm. Overall, the MAE,PSNR and NCC were 60.9±11.7 HU, 24.6±0.9 dB, and 0.96±0.01. We have developed a novel deep learning-basedmethod to generate sCT with a high accuracy. The proposed method makes the sCT comparable to that of the planningCT. With further evaluation and clinical implementation, this method could be a useful tool for MRI-based radiationtreatment planning and attenuation correction in a PET/MRI scanner.
展开▼
