3059 - Scalable Implementation of CBCT-Guided Adaptive Radiotherapy: Impact of RTT Integration, Advanced CBCT and AI Contouring
Presenter(s)
M. H. Lin1, D. D. M. Parsons1, D. J. Sher1, and S. N. Badiyan2; 1Department of Radiation Oncology, University of Texas Southwestern Medical Center, Dallas, TX, 2University of Texas Southwestern Medical Center, Department of Radiation Oncology, Dallas, TX
Purpose/Objective(s): To evaluate the longitudinal impact of RTT-led OAR editing, strategic case selection, imaging enhancement, and AI contour integration on CBCT-guided adaptive radiotherapy (ART) workflow efficiency and scalability.
Materials/Methods: A longitudinal analysis (2021–2025) was performed across eight treatment sites. Machine activity log files were extracted and parsed to quantify step-level workflow metrics, including influencer contour time, target contour time, plan generation and review time, and adaptive workflow (AWF) time. Influencer and target contour times were combined to generate total contour time. Weighted yearly means were calculated to assess workflow evolution across sequential clinical and technological phases. The program transitioned from physician-led contouring (2021) to an RTT-integrated model in which RTTs performed influencer and daily OAR edits prior to physician review (mid-2022). Subsequent phases included expansion to higher-complexity adaptive indications (2023), deployment of advanced CBCT imaging (Jan. 2024), and integration of whole-body AI contouring (Jan. 2025).
Results:
A total of 9,465 adaptive sessions were analyzed. Case mix evolved substantially, reflecting deliberate strategic redistribution rather than technical limitation. After RTT-led OAR editing implementation, annual adaptive cases increased from 1670 to 3330 . Although physician editing burden decreased and scheduling flexibility improved, total contour time increased from 7.0 ± 3.7 minutes (2021) to 10.4 ± 4.1 minutes (2022), reflecting incorporation of physician review latency. With expansion to more complex cases, total contour time rose to 12.8 ± 2.2 minutes (2023) and peaked at 14.1 ± 2.6 minutes (2024); AWF increased from 20.4 ± 5.8 minutes (2022) to 23.9 ± 4.6 minutes (2024). Following advanced CBCT deployment, abdomen adaptive utilization increased 51% year-over-year in 2024 and an additional 30% in 2025 (+96% vs 2023). After AI contour integration, total contour time decreased to 11.6 ± 3.5 minutes and AWF to 22.8 ± 4.8 minutes despite continued case growth. Breast demonstrated the largest contour reduction (-9.2 minutes).Conclusion: ART scalability was driven primarily by workflow redesign and strategic deployment rather than platform limitation. Task redistribution enabled volume expansion, imaging enhanced adoption, and AI integration restored efficiency, supporting durable and scalable clinical implementation.
Table 1. Intervention and Efficiency (min)| Year | Intervention | ART Tx# | Total Contour Time (Influencer + Target) | Plan Generation & Review | Adaptive Workflow (AWF) |
| 2021 | Physician-led process (baseline) | 110 | 7.0 ± 3.7 | 5.8 ± 1.7 | 16.4 ± 5.7 |
| 2022 | RTT edits OAR | 1670 | 10.4 ± 4.1 | 5.7 ± 1.1 | 20.4 ± 5.8 |
| 2023 | Complexity expansion | 2098 | 12.8 ± 2.2 | 6.0 ± 1.1 | 22.1 ± 3.2 |
| 2024 | Advanced CBCT | 2437 | 14.1 ± 2.6 | 7.4 ± 2.2 | 23.9 ± 4.6 |
| 2025 | Whole-body AI contour | 3330 | 11.6 ± 3.5 | 8.7 ± 1.3 | 22.8 ± 4.8 |