Brooks, Jamison

Abstract Background Deep learning auto segmentation (DLAS) tools are widely used for radiation therapy planning, yet limited information exists on how patient positioning impacts their performance, particularly in head and neck (HN) computed tomography (CT) imaging. Purpose This study investigates the impact of abnormal neck tilt on the accuracy of DLAS‐generated contours in HN CT scans. We hypothesize that abnormal positioning degrades auto segmentation performance, particularly for organs at risk (OARs) influenced by cervical spine orientation. Methods A total of 35 HN CT scans were retrospectively analyzed. Neck tilt was quantified using principal component analysis of the spinal cord contours from C1 to C4, with patients stratified into normal and abnormal tilt groups based on percentile thresholds. Seven FDA‐cleared DLAS tools were evaluated for OAR segmentation accuracy across brainstem, parotid glands, submandibular glands, brachial plexus, and optic nerves. Gold‐standard contours were curated from manually created clinical contours and compared to DLAS output using Sorenson's volumetric dice similarity coefficient (DSC), surface Dice similarity coefficient with a 2 mm threshold (sDSC), mean distance to agreement (MDA) and difference in mean dose. Statistical differences were assessed using Wilcoxon rank‐sum tests ( p < 0.05). Results Among the evaluated OARs, the parotid glands showed consistent and statistically significant degradation in contouring accuracy for patients with abnormal neck tilt across all metrics (DSC, sDSC, MDA) and across all seven DLAS tools. Changes in contouring accuracy resulted in significant differences in both absolute and signed difference in mean dose. No consistent differences were observed for other structures across multiple DLAS tools. Conclusions Abnormal neck tilt is associated with reduced DLAS accuracy and greater dose variability for parotid gland segmentation in HN radiotherapy planning. These findings underscore the need for enhanced patient quality assurance strategies for patients with abnormal positioning when using clinical DLAS workflows. Future work should explore automated detection of anatomical outliers and site‐specific model retraining to ensure accurate contouring delineation for patients with atypical positioning or anatomy.

Abstract Objective . An algorithm was developed for automated positioning of lattice points within volumetric modulated arc lattice radiation therapy (VMAT LRT) planning. These points are strategically placed within the gross tumor volume (GTV) to receive high doses, adhering to specific separation rules from adjacent organs at risk (OARs). The study goals included enhancing planning safety, consistency, and efficiency while emulating human performance. Approach . A Monte Carlo-based algorithm was designed to optimize the number and arrangement of lattice points within the GTV while considering placement constraints and objectives. These constraints encompassed minimum spacing between points, distance from OARs, and longitudinal separation along the z -axis. Additionally, the algorithm included an objective to permit, at the user’s discretion, solutions with more centrally placed lattice points within the GTV. To validate its effectiveness, the automated approach was compared with manually planned treatments for 24 previous patients. Prior to clinical implementation, a failure mode and effects analysis (FMEA) was conducted to identify potential shortcomings. Main results. The automated program successfully met all placement constraints with an average execution time (over 24 plans) of 0.29 ±0.07 min per lattice point. The average lattice point density (# points per 100 c.c. of GTV) was similar for automated (0.725) compared to manual placement (0.704). The dosimetric differences between the automated and manual plans were minimal, with statistically significant differences in certain metrics like minimum dose (1.9% versus 1.4%), D5% (52.8% versus 49.4%), D95% (7.1% versus 6.2%), and Body-GTV V30% (20.7 c.c. versus 19.7 c.c.). Significance. This study underscores the feasibility of employing a straightforward Monte Carlo-based algorithm to automate the creation of spherical target structures for VMAT LRT planning. The automated method yields similar dose metrics, enhances inter-planner consistency for larger targets, and requires fewer resources and less time compared to manual placement. This approach holds promise for standardizing treatment planning in prospective patient trials and facilitating its adoption across centers seeking to implement VMAT LRT techniques.