Inter-fractional magnitude and trajectory changes are of great importance for radiotherapy (RT) of movingtargets. In order to verify the amount and characteristics of patient-specific respiratory motion prior to eachRT treatment session, a time-resolved cone-beam computed tomography (4D CBCT) is necessary. However, dueto sparse view artifacts, the resulting image quality is limited when applying current 4D CBCT reconstructionapproaches. In this study, a new deep learning-based boosting approach for 4D CBCT reconstruction is presentedthat does not rely on any a-priori information (e.g. 4D CT images) and is applicable to arbitrary reconstructionalgorithms. It is shown that the overall image quality is significantly improved after boosting; in particular,sparse view sampling artifacts are suppressed.
展开▼
