Abstract
The goal of this work is to evaluate the dosimetric impact of an overshooting phenomenon in step-and-shoot IMRT delivery, and to demonstrate a novel method to mitigate the issue. Five pelvis IMRT patients treated on Varian 2100C EX linacs with larger than +4.5% phantom ion chamber point-dose difference relative to planned dose were investigated. For each patient plan, 5 fractions were delivered. DynaLog files were recorded and centi-MU pulses from dose integrator board for every control point (CP) were counted using a commercial pulse counter. The counter recorded CP MU agrees with DynaLog records, both showing an ~ 0.6MU overshoot of the first segment of every beam. The 3D patient dose was recalculated from the counter records and compared to the planned dose, showing that the overshoot resulted in on average 2.05% of PTV D95 error, and 2.49%, 2.61% and 2.45% of D1cc error for rectum, bladder, and bowel, respectively. The initial plans were then modified by inserting a specially designed MLC segment to the start of every beam. The modified plans were also delivered five times. The dose from the modified delivery was calculated using counter recorded CP MU. The corresponding Dx parameters were all within 0.31% from the original plan. IMRT QA results also show a 2.2% improvement in ion chamber point-dose agreement. The results demonstrate that the proposed plan modification method effectively eliminates the overdosage from the overshooting phenomenon.