Evaluation of efficacy in prostate cancer contouring for MR-guided radiotherapy using structured contouring training modules

Purpose Magnetic resonance (MR)-guided radiotherapy offers superior soft-tissue visualization, enabling more precise contouring and reduced toxicity. However, this workflow poses challenges for trainees. This study evaluated whether structured contouring training modules improve contouring accuracy and efficiency for radiation oncology trainees. Methods Residents contoured the prostate and pelvic organs-at-risk (OARs) from MR images in three phases: baseline, after reviewing a written contouring module, and after receiving personalized feedback with a training video. Automated tools using Boolean logic provided quantitative and visual feedback. Resident contours were compared against gold-standard contours established by staff radiation oncologists. Accuracy was assessed using Dice Similarity Coefficient (DSC), 95th percentile Hausdorff Distance (HD), and Absolute Volume Difference (AVD). Efficiency was measured as contouring time per case. Friedman and Wilcoxon signed-rank tests with Benjamini–Hochberg adjustment were applied, and significance was defined as p < 0.05. Results Four trainees contoured five cases in each phase. When data were combined across structures, HD and AVD improved significantly (8.332 mm to 6.117 mm, p = 0.0498; 6.070 mL to 2.695 mL, p = 0.039), along with contouring time (40 min to 24 min, p = 0.018). DSC increased from 0.875 to 0.910 ( p = 0.174). Structure-specific analysis showed significant improvement in at least one metric for all structures except the seminal vesicles. Conclusion Structured MR contouring training improved both accuracy and efficiency. Incorporating this training into residency curricula may better prepare residents for MR-guided radiotherapy and make adaptive radiotherapy workflows more streamlined.