Clinical Implementation and Evaluation of an Artificial Intelligence-Driven One-Click Automatic Planning System for Functional Lung Avoidance Radiation Therapy
Related Experiment Video
Updated: Sep 18, 2026
Radiation Planning Assistant - A Web-based Tool to Support High-quality Radiotherapy in Clinics with Limited Resources
Published on: October 6, 2023
Tianyu Xiong1, Zhi Chen2, Chengcheng Fan3
1Department of Radiation Oncology, Sun Yat-Sen University Cancer Center, State Key Laboratory of Oncology in South China, Guangdong Provincial Clinical Research Center for Cancer, Guangzhou, People's Republic of China; Department of Health Technology and Informatics, The Hong Kong Polytechnic University, Hong Kong SAR, People's Republic of China.
Purpose:
This study aimed to advance the clinical application of an artificial intelligence-driven automatic planning system for functional lung avoidance radiation therapy (AP-FLART) through clinical implementation and comprehensive evaluation.
Methods And Materials:
AP-FLART integrates dosimetric score-based beam angle selection, multimodality-guided dose prediction, and function-guided dose mimicking. The system was enhanced and implemented within the clinical treatment planning system, RayStation. Clinical performance and potential benefits were assessed using a test data set comprising 33 patients with lung cancer who underwent Single-Photon Emission Computed Tomography (SPECT) ventilation or perfusion imaging and lung radiation therapy. Dosimetric metrics and normal tissue complication probabilities of automatic FLART plans were compared with manual conventional radiation therapy and FLART plans created by an experienced planner. Three clinicians conducted blinded reviews comparing the automatic and manual-FLART plans.
Results:
Compared with manual conventional radiation therapy plans, automatic FLART plans significantly reduced the high-function lung mean dose by 15.1%. Among FLART-benefiting patients, automatic FLART plans reduced the probability of grade ≥2 radiation pneumonitis by 6.25 percentage points (27%), while maintaining comparable probabilities for other side effects. The estimated clinical benefits are similar to those of manual-FLART plans. Blinded reviews indicated that 87.9% of automatic FLART plans were clinically acceptable without modification, and 68.7% were rated as comparable (38.4%) or superior (30.3%) to manual-FLART plans. Furthermore, AP-FLART reduced planning time from 2 to 3 hours for manual-FLART planning to approximately 8 minutes.
Conclusions:
A clinically viable, one-click AP-FLART has been successfully implemented and comprehensively evaluated. AP-FLART shows considerable promise for improving the efficiency and reducing workload demands of FLART planning and advancing the broader clinical adoption of FLART.
