Scatter Artifacts Removal Using Learning-Based Method for CBCT in IGRT System

Abstract

Cone-beam-computed tomography (CBCT) has shown enormous potential in recent years, but it is limited by severe scatter artifacts. This paper proposes a scatter-correction algorithm based on a deep convolutional neural network to reduce artifacts for CBCT in an image-guided radiation therapy (IGRT) system. A two-step registration method that is essential in our algorithm is implemented to preprocess data before training. The testing result on real data acquired from the IGRT system demonstrates the ability of our approach to learn artifacts distribution. Furthermore, the proposed method can be applied to enhance the performance on such applications as dose estimation and segmentation.