ESTRO 2025 - Abstract Book

S1202

Clinical – Lower GI

ESTRO 2025

2290

Digital Poster First implementation of an All-in-One solution with full automatic workflow during for rectal cancer based on an integrated CT-linac Luqi Wang 1 , Ran Peng 1 , Xuemin Li 1 , Mingqing Wang 1 , Siyi Lu 1 , Wei Zhang 2 , Lecheng Jia 3 , Mengying Yang 4 , Hao Wang 1,5 1 Radiation Oncology, Peking University 3rd Hospital, Beijing, China. 2 Radiotherapy Business Unit, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, China. 3 Innovative Medical Equipment, United Imaging Research Institute, Shenzhen, China. 4 Intelligent Imaging, United Imaging Research Institute, Beijing, China. 5 Cancer Center, Peking University 3rd Hospital, Beijing, China Purpose/Objective: To delineate the technical characteristics and clinical performance of an All-in-One solution with full-workflow designed for patients with rectal cancer based on a computed tomography (CT)-linear accelerator(linac) combination. Material/Methods: The "All-in-One" workflow, which is predicated on a CT-linac combination and the application of artificial intelligence models, includes simulation, auto-segmentation, auto-planning, image guidance, beam delivery, and in vivo quality assurance (QA). This study involved 30 patients diagnosed with T 3-4 stage rectal cancer, treated at Peking University Third Hospital between May and September 2024. The time-cost of each procedure, the outcomes of auto segmentation and auto-planning, the real-time passing rate of γ ray, and the change of bladder volume based on the fan-beam CT images were collected and discussed in this research. Results: For the patients included in the study, it was essential to perform manual adjustments to the auto-segmented target volumes. The duration of the whole process was 38.8 ± 6.9 minutes for the enrolled patients, depending mostly on the time required for auto-segmentation and verification, which takes an average of 1121.2 seconds. The Dice similarity coefficient of CTV45, CTV50, and GTV before and after the modifications were 0.93±0.03, 0,88±0.07, and 0.81±0.13. The auto-segmented normal tissues were clinically acceptable without any modifications. Only one case required manual adjustment of the automatic plan. The pretreatment image-guided radiation therapy (IGRT) corrections were within 3 mm in all directions, and the EPID-based in vivo QA showed pass rate of γ ray (3D) above 95% (3%/3 mm/10% threshold). The alterations in bladder volume observed in these 30 patients exhibited two distinct clustered distributions.