Evaluation of a commercial automatic treatment planning system for liver stereotactic body radiation therapy treatments

PURPOSE: Automated treatment planning is a new frontier in radiotherapy. The Auto-Planning module of the Pinnacle3 treatment planning system (TPS) was evaluated for liver stereotactic body radiation therapy treatments. METHODS: Ten cases were included in the study. Six plans were generated for each case by four medical physics experts. The first two planned with Pinnacle TPS, both with manual module (MP) and Auto-Planning one (AP). The other two physicists generated two plans with Monaco TPS (VM). Treatment plan comparisons were then carried on the various dosimetric parameters of target and organs at risk, monitor units, number of segments, plan complexity metrics and human resource planning time. The user dependency of Auto-Planning was also tested and the plans were evaluated by a trained physician. RESULTS: Statistically significant differences (Anova test) were observed for spinal cord doses, plan average beam irregularity, number of segments, monitor units and human planning time. The Fisher-Hayter test applied to these parameters showed significant statistical differences between AP e MP for spinal cord doses and human planning time; between MP and VM for monitor units, number of segments and plan irregularity; for all those between AP and VM. The two plans created by different planners with AP were similar to each other. CONCLUSIONS: The plans created with Auto-Planning were comparable to the manually generated plans. The time saved in planning enables the planner to commit more resources to more complex cases. The independence of the planner enables to standardize plan quality.