Unencrypted login | home

Program Information

4D MIMVista Workflow Vs 4D Dose Calculations Based On a Monte Carlo Method


G Valdes

G Valdes1*, K Iwamoto2, P Lee3, D Low4, C Robinson5, J Lamb6, (1) UCLA, Los Angeles, CA, (2) UCLA, Los Angeles, California, (3) UCLA, Los Angeles, California, (4) UCLA, Los Angeles, CA, (5) Washington University in St. Louis, Louis,, MO, (6) University of California, Los Angeles, Los Angeles, CA

SU-E-T-435 Sunday 3:00PM - 6:00PM Room: Exhibit Hall

Purpose: We investigated the accuracy of a commercially-available software algorithm for 4D dose accumulation that consists of rigidly projecting the 3D dose calculated in the reference respiratory phase onto the remaining phases, followed by deformable registration, versus 4D dose calculation involving Monte Carlo recalculations of the original treatment plan applied to each of the phases.

Methods:Radiotherapy plans for stage I-II lung cancer were created for 8 patients who had respiratory-correlated computed tomography (4D-CT) performed as part of an IRB-approved research protocol. Plans were generated and evaluated using the dose calculated by a Monte Carlo algorithm on the end-exhale 4D-CT phase. 3D doses for all other phases and the 4D dose were calculated using a rigid dose-projection workflow as part of a commercially-available software package. The resulting DVHs were compared with those obtained using Monte Carlo calculations after the initial plans were recalculated on each of the respiratory phases. In both cases the final dose accumulation was performed using the same commercially available deformable registration algorithm. The calculated dose distributions were compared according to the principal dosimetric indices used in our clinic to evaluate gross tumor volume (GTV) coverage: dose received by 99% of the gross tumor volume (D99(GTV)), minimum dose received by the GTV (Dmin(GTV)), and mean dose received by the GTV (Dmean(GTV)).

Results:Mean differences between the commercial dose projection and the Monte Carlo recalculation algorithms were 0.1 Gy, 0.3 Gy, and 0.2 Gy for D99(GTV), Dmin(GTV) and Dmean(GTV). Maximum differences were 2.1 Gy, 2.9 Gy, and 2.2 Gy, respectively, between the two methods.

Conclusion:The commercial dose accumulation workflow using rigid dose projection is much faster than 4D accumulation with Monte Carlo dose recalculation, and no large dosimetric differences were found between the methods. This study lends support to its use in a clinical setting.

Contact Email: