Knowledge Based Planning Shortens Planning Time for Diffuse Midline Glioma
Abstract
Purpose
Diffuse midline glioma (DMG) is an aggressive central nervous system malignancy typically affecting children and requiring urgent radiotherapy treatment. This study evaluates the application of knowledge-based planning to expedite treatment planning and treatment start, while maintaining plan quality for paediatric patients with brainstem diffuse midline glioma.
Methods
Ten consecutive paediatric patients (ages 3-15) treated over a four-year period were retrospectively analysed. A RapidPlan (RP) model was built using 21 historical cases. Two radiation therapists, one being a paediatric specialist and the other a non-specialist, created conventional VMAT plans and also RP-generated plans for comparison. Planning time, target coverage, conformity index (CI95%), and doses to organs at risk were evaluated.
Results
The average planning time for RP was significantly shorter at 17 minutes (range 9-28) compared to 47 minutes (range 22-71) for conventional VMAT planning, i.e. a reduction of 64%. All plans achieved clinically acceptable coverage, with mean PTV coverage of 98.1% for VMAT versus 97.2% for RP. RP plans demonstrated superior conformity (mean CI95% 1.05 vs 1.24 for VMAT; p=0.00004). No significant differences were found in doses to organs at risk, though a trend toward lower pituitary doses was observed with RP plans (35.4Gy vs 39.4Gy).
Conclusion
Knowledge-based planning using RapidPlan significantly reduced treatment planning time for paediatric DMG while maintaining plan quality and improving conformity. This finding was demonstrated for planners with specialist paediatric planning expertise and also for those with less paediatric experience. RapidPlan enables efficient high‑quality VMAT planning, supporting same‑day plan‑and‑treat workflows that are essential for DMG cases requiring emergency treatment.