Background: Organs at risk (OARs) delineation is a crucial step of radiotherapy (RT) treatment planning workflow. Time-consuming and inter-observer variability are main issues in manual OAR delineation, mainly in the head and neck (H & N) district. Deep-learning based auto-segmentation is a promising strategy to improve OARs contouring in radiotherapy departments. A comparison of deep-learning-generated auto-contours (AC) with manual contours (MC) was performed by three expert radiation oncologists from a single center. Methods: Planning computed tomography (CT) scans of patients undergoing RT treatments for H&N cancers were considered. CT scans were processed by Limbus Contour auto-segmentation software, a commercial deep-learning auto-segmentation based software to generate AC. H&N protocol was used to perform AC, with the structure set consisting of bilateral brachial plexus, brain, brainstem, bilateral cochlea, pharyngeal constrictors, eye globes, bilateral lens, mandible, optic chiasm, bilateral optic nerves, oral cavity, bilateral parotids, spinal cord, bilateral submandibular glands, lips and thyroid. Manual revision of OARs was performed according to international consensus guidelines. The AC and MC were compared using the Dice similarity coefficient (DSC) and 95% Hausdorff distance transform (DT). Results: A total of 274 contours obtained by processing CT scans were included in the analysis. The highest values of DSC were obtained for the brain (DSC 1.00), left and right eye globes and the mandible (DSC 0.98). The structures with greater MC editing were optic chiasm, optic nerves and cochleae. Conclusions: In this preliminary analysis, deep-learning auto-segmentation seems to provide acceptable H&N OAR delineations. For less accurate organs, AC could be considered a starting point for review and manual adjustment. Our results suggest that AC could become a useful time-saving tool to optimize workload and resources in RT departments.

Background: The aim of this study is to evaluate the delta radiomics approach based on mesorectal radiomic features to develop a model for predicting pathological complete response (pCR) and 2-year disease-free survival (2yDFS) in locally advanced rectal cancer (LARC) patients undergoing neoadjuvant chemoradiotherapy (nCRT). Methods: Pre- and post-nCRT MRIs of LARC patients treated at a single institution from May 2008 to November 2016 were retrospectively collected. Radiomic features were extracted from the GTV and mesorectum. The Wilcoxon–Mann–Whitney test and area under the receiver operating characteristic curve (AUC) were used to evaluate the performance of the features in predicting pCR and 2yDFS. Results: Out of 203 LARC patients, a total of 565 variables were evaluated. The best performing pCR prediction model was based on two GTV features with an AUC of 0.80 in the training set and 0.69 in the validation set. The best performing 2yDFS prediction model was based on one GTV and two mesorectal features with an AUC of 0.79 in the training set and 0.70 in the validation set. Conclusions: The results of this study suggest a possible role for delta radiomics based on mesorectal features in the prediction of 2yDFS in patients with LARC.

Introduction: The main goal of radiotherapy (RT) is to deliver a precise dose to the target while sparing the surrounding normal tissue and minimizing side effects. Appropriate patient immobilization is crucial, especially for head and neck cancer (HNC) and Brain Cancer (BC). Conventional closed-face masks (CFMs), while effective in minimizing head motion, can cause significant discomfort, anxiety, and claustrophobia. Open-face masks (OFMs) have been developed to increase patient comfort while providing precise immobilization. Methods: Following the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) extension for scoping reviews and the Arskey and O’Malley framework, an electronic search of EMBASE, PubMed, SCOPUS, and Web of Science was conducted to identify original studies reporting the use and description of OFMs in clinical practice up to April 2024. The inclusion criteria were English-language articles focusing on OFMs for HNC and BC patients undergoing RT. Results: Of 618 titles, 19 articles fulfilled the selection criteria. Most studies were comparative (n = 13) or observational (n = 6). The articles were categorized by treatment site, resulting in three groups: BC (n = 14, 68.4%), HNC (n = 4, 21.4%), and mixed (n = 2, 10.5%), which includes both BC and HNC. Of note, 82.4% (n = 16) of the included studies were published from 2020 onwards, emphasizing the recent adoption of OFM in clinical practice. Conclusions: The reviewed studies show that OFMs, in combination with SGRT, offer significant advantages in terms of patient comfort and positioning accuracy in HNC and BC treatments. Reproducibility in the sub-millimeter and sub-degree range can be achieved, which supports the use of OFMs in clinical practice. Future research should explore innovative combinations of immobilization and monitoring to further improve RT outcomes and ensure precise treatment while increasing patient comfort.

Recent advances in Radiotherapy have led to the development of online adaptive RT (oART), a procedure addressing inter-fraction anatomical variations. Integrating artificial intelligence (AI) into the oART procedure speeds up the process and reduces user dependency. This study investigates the dosimetric advantage of implementing AI-driven oART in prostate cancer. A total of 31 prostate cancer patients treated with oART on an AI-integrated Linac were analyzed. Patients were categorized by nodal involvement. For prostate-only cases, the Clinical Target Volume (CTV) included the prostate and seminal vesicles (CTV1), with a 5 mm margin (8 mm caudally) for Planning Target Volume (PTV), named PTV1. For nodal cases, pelvic lymph nodes were added (and categorized as CTV2) with a 5 mm isotropic margin (PTV2). Daily CBCTs were acquired, with OARs (rectum, bladder, bowels) automatically segmented by the AI system, while targets were manually delineated. Two plans were generated: a predicted one, calculating the original plan's fluence on daily anatomy, and an adapted one, with complete fluence re-optimization. Daily DVH indicators for PTV(V95%), CTV(D98%), bladder (V65Gy), bowel (V45Gy), and rectum (V50Gy) were compared between predicted and adapted plans using the Wilcoxon-Mann-Whitney test. Total session time, from CBCT acquisition to treatment completion, was also recorded. oART treatment improved prostate coverage in both patient groups (+10.4% and +11.8% in PTV V95% for patients with and without lymph nodes) and CTV D98% (+2.6% with lymph nodes, +2.9% without). Improvements for arm 2 were smaller (+3.1% in PTV2 V95%, +2.2% in CTV2 D98%). Statistical differences were insignificant in OAR DVH indicators (p > 0.1). Median treatment time was 25 min and 32 min for prostate-only and lymph node cases, respectively. This study demonstrates that oART in prostate cancer results in a significant improvement in target coverage with no significant difference in OARs.

**Abstract text** Purpose: The utilization of artificial intelligence (AI) and natural language processing (NLP) methodologies have demonstrated | POSTER