Experience of Implementing Deep Learning–Based Automatic Contouring in Breast Radiation Therapy Planning: Insights From Over 2000 Cases

Abstract

Background and purposeThis study evaluates the impact and clinical utility of an auto-contouring system for radiation therapy treatments. Materials and methodsThe auto-contouring system was implemented in 2019. We evaluated data from 2,428 patients who underwent adjuvant breast radiation therapy before and after the system's introduction. We collected the treatment's finalized contours, reviewed and revised by multidisciplinary team. After implementation, the treatment contours underwent a finalization process which involved manual review and adjustment of the initial auto-contours. For the pre-implementation group (n = 369), auto-contours were generated retrospectively. We compared the auto-contours and final contours using the Dice similarity coefficient (DSC) and the 95% Hausdorff distance (HD95). ResultsWe analyzed 22,215 structures from final and corresponding auto-contours. The final contours were generally larger, encompassing more slices in the superior or inferior directions. For organs-at-risk (OAR), the heart, esophagus, spinal cord, and contralateral breast demonstrated significantly increased DSC and reduced HD95 post-implementation (all p < 0.05), except for the lungs, which presented inaccurate segmentation. Among target volumes, CTVn_L2, L3, L4, and IMN showed increased DSC and reduced HD95 in the post-implementation group (all p < 0.05), although the increase was less pronounced than OAR outcomes. The analysis also covered factors contributing to significant differences, pattern identification, and outlier detection. ConclusionsIn our study, the adoption of the auto-contouring system was associated with an increased reliance on automated settings, underscoring its utility and the potential risk of automation bias. Given these findings, we believe it may be important to consider the integration of stringent risk assessments and quality management strategies as a precautionary measure for the optimal use of such systems.