Real-world validation of Artificial Intelligence-based Computed Tomography auto-contouring for prostate cancer radiotherapy planning

• Published in: Physics and imaging in radiation oncology 2023-10, Vol.28, p.100501-100501, Article 100501
• Main Authors: Palazzo, Gabriele, Mangili, Paola, Deantoni, Chiara, Fodor, Andrei, Broggi, Sara, Castriconi, Roberta, Ubeira Gabellini, Maria Giulia, del Vecchio, Antonella, Di Muzio, Nadia G., Fiorino, Claudio
• Format: Article
• Language: English
• Subjects: Artificial Intelligence in radiotherapy; Auto-contouring; Original; Planning; Prostate cancer; Radiotherapy
• ISSN: 2405-6316; 2405-6316
• DOI: 10.1016/j.phro.2023.100501

•Validation of commercial auto-contouring software for prostate cancer planning.•Variability of manual vs automatic contouring is lower than inter-observer variability.•Quick editing further improved performances, reducing inter-observer variability.•Contouring time was reduced from 17–24 min to 3–7 min.•An optimized and integrated automatic workflow was clinically implemented. Artificial Intelligence (AI)-based auto-contouring for treatment planning in radiotherapy needs extensive clinical validation, including the impact of editing after automatic segmentation. The aims of this study were to assess the performance of a commercial system for Clinical Target Volumes (CTVs) (prostate/seminal vesicles) and selected Organs at Risk (OARs) (rectum/bladder/femoral heads + femurs), evaluating also inter-observer variability (manual vs automatic + editing) and the reduction of contouring time. Two expert observers contoured CTVs/OARs of 20 patients in our Treatment Planning System (TPS). Computed Tomography (CT) images were sent to the automatic contouring workstation: automatic contours were generated and sent back to TPS, where observers could edit them if necessary. Inter- and intra-observer consistency was estimated using Dice Similarity Coefficients (DSC). Radiation oncologists were also asked to score the quality of automatic contours, ranging from 1 (complete re-contouring) to 5 (no editing). Contouring times (manual vs automatic + edit) were compared. DSCs (manual vs automatic only) were consistent with inter-observer variability (between 0.65 for seminal vesicles and 0.94 for bladder); editing further improved performances (range: 0.76–0.94). The median clinical score was 4 (little editing) and it was