Pseudo-medical image-guided technology based on 'CBCT-only' mode in esophageal cancer radiotherapy

•A new IGRT technique based on the "CBCT-only" mode is proposed to improve the treatment effectiveness of esophageal cancer radiotherapy. This mode relies on CBCT images and utilizes different deep learning models to synthesize corresponding pseudo-CT and pseudo-PET images. The synthesized data is used to assist clinicians in tasks such as delineation, dose calculation, and positioning verification during radiotherapy, thus avoiding various sources of human error introduced in current clinical radiotherapy practices.•A dual-domain parallel deep learning model with attention waning mechanism (AWM-PNet) is proposed to obtain pseudo-CT images from CBCT. The parallel image synthesis approach can effectively leverage the information from sinogram and spatial domains. The attention waning mechanism captures important features and contextual information in a more efficient manner.•A GAN model with a prior region aware mechanism (PRAM-GAN) is employed to synthesize pseudo-PET images from CT. By incorporating knowledge from clinical experts regarding tumor location and shape, the mechanism assists the model to learn the representation of tumor region features in the PET image domain and their relationship with other soft tissues.•The accuracy of the pseudo-medical images synthesized by the models proposed in this study has been verified in terms of anatomical representation and dosimetric evaluation. Purpose To minimize the various errors introduced by image-guided radiotherapy (IGRT) in the application of esophageal cancer treatment, this study proposes a novel technique based on the 'CBCT-only' mode of pseudo-medical image guidance. Methods The framework of this technology consists of two pseudo-medical image synthesis models in the CBCT→CT and the CT→PET direction. The former utilizes a dual-domain parallel deep learning model called AWM-PNet, which incorporates attention waning mechanisms. This model effectively suppresses artifacts in CBCT images in both the sinogram and spatial domains while efficiently capturing important image features and contextual information. The latter leverages tumor location and shape information provided by clinical experts. It introduces a PRAM-GAN model based on a prior region aware mechanism to establish a non-linear mapping relationship between CT and PET image domains. As a result, it enables the generation of pseudo-PET images that meet the clinical requirements for radiotherapy. Results The NRMSE and multi-scale SSIM (MS-SSI