Purpose: Cone-beam computed tomography (CBCT) is frequently used for accurate image guided radiation therapy (IGRT). However, the poor CBCT image quality prevents its further clinical use. Thus, it is important to improve the HU accuracy and structure preservation of CBCT images.
Methods: In this study, we proposed a novel method to generate synthetic CT (sCT) images from CBCT images. A multi-resolution residual deep neural network (RDNN) was adopted for image regression from CBCT images to planning CT (pCT) images. At the coarse level, RDNN was first trained with a large amount of lower resolution images, which can make the network focus on coarse information and prevent overfitting problems. More fine information was obtained gradually by fine-tuning the coarse model using fewer number of higher resolution images. Our model was optimized by using aligned pCT and CBCT image pairs of a particular body region of 153 prostate cancer patients treated in our hospital (120 for training, 33 for testing).
Results: The averaged accuracy of mean absolute error (MAE) between CBCT and pCT on testing data was 121.61 HU, while the MAE between the sCT and pCT images was only 27.62 HU. And the averaged structural similarity index measure between sCT and CBCT was 48.89% higher than that of CBCT and pCT. Our proposed multi-resolution RDNN also provides an overall better image quality in terms of HU accuracy and structural fidelity than U-Net and RDNN trained only using high-resolution CBCT and pCT image pairs.
Conclusion: The sCT images generated by our proposed multi-resolution RDNN have higher HU accuracy and structural fidelity, which may promote the further applications of CBCT images in the clinic for structure segmentation, dose calculation and adaptive radiotherapy planning.