Iterative development and technical framework of the AI-based automated planning model. (IMAGE)

Beijing Institute of Technology Press Co., Ltd

Iterative development and technical framework of the AI-based automated planning model.

Caption

(A) Overview of the full pipeline for online AIO radiotherapy planning, from clinical prescription to deliverable plan. (B) Baseline model (V1): A CAD-3D U-Net-based dose prediction module combined with multidimensional plan adjustment. (C) V2 upgrades: introduction of label-guided pareto-optimal dose selection and a priority-based constraint mechanism to balance target coverage and OAR sparing. PTV, planning target volume. (D) V3 upgrades: improvements in robustness for complex T4 cases via quantile loss, dataset enrichment, and stochastic platform optimization. MSE, mean squared error. (E) V4 upgrades: computational acceleration through CPU-parallelized optimization, Monte Carlo dose learning, and GPU-based final dose calculation.

Credit

Ying Sun, State Key Laboratory of Oncology in South China, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Guangdong Provincial Clinical Research Center for Cancer, Sun Yat-sen University Cancer Center.

Usage Restrictions

News organizations may use or redistribute this image, with proper attribution, as part of news coverage of this paper only.

License

Licensed content

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.