SU-E-T-724: Testing the Limitations of a Photon Dose Algorithm in Commissioning.

PURPOSE: This work is to test the limitations of a photon dose algorithm during the commissioning of a treatment planning system.

METHODS: Commissioning tests for a model-based dose calculation algorithm typically involve the verification of PDD, output and dose profiles, but often those tests are performed at one SSD, one depth and one nominal gantry position. Those tests are useful to validate the modeling of radiation beams, but they may not be sufficient to detect the limitations of an algorithm. Additional tests were performed at several SSDs, various depths, oblique gantry angles and off-axis conditions to demonstrate the accuracy and limitations of the dose calculation algorithm. Dose profiles were measured by in-water scanning with a three-dimensional scanning system and calculated in the TPS using a virtual water phantom.

RESULTS: Dose profiles from both the measurement and the calculation were plotted for comparison. Also plotted were the difference in absolute dose between the measurement and the calculation. For the off-axis MLC field tests, larger discrepancies were observed at the open portion and the far edge of the field, which became greater with short SSD and deep depth. In addition, the shape of the profile appeared different for the open portion of the field. For the oblique MLC field tests, the largest difference was seen around the high-dose shoulder of the profile. Differences in some point doses were observed to be greater than 10% of the maximum field dose.

CONCLUSION: We have learned the limitations of the superposition dose algorithm in the XiO planning system through some specific tests, which are related to the parallel kernel approximation and the simplified energy spectra employed by the algorithm to calculate field-size dependent fluence.