Analysing collimator structure effects in head-scatter calculationsfor IMRT class fields using scatter raytracing

Abstract

The frequent blocking of the irradiated volume in intensity modulatedradiation therapy (IMRT) makes the head-scatter fraction of the incidentphoton fluence more significant than that in conventional therapy with openfields. On the other hand, certain collimator configurations block scatterphotons directed to a given observation point while allowing primary photonsto be transmitted. The `anomalous blocking' makes the primary field a poorindicator of the scatter fluence. Since large MU-to-cGy ratios in IMRT canmagnify head-scatter uncertainties, it becomes necessary to accurately modelboth the effective scatter source and the collimator structure that limits thescatter reaching the irradiated volume. First we obtain a dual-source model,using a Taylor series expansion to derive the effective scatter sourcedistribution from the data measured for the Elekta SL20 linac equipped with amulti-leaf collimator (MLC). Then, using a raytracing algorithm, we calculatethe transmission of scatter rays from the effective scatter source plane topoints in the patient plane. The method can account for the anomalous blockingof scatter by the MLC leaves and the back-up diaphragms. For a variety ofcollimator settings tested, the calculations agree with measurements to anaccuracy of 0.002Ψ10×10, where Ψ10×10 is the total(primary + scatter) photon fluence of an open 10×10 cm2 field forthe same MU delivered. Although the significance of collimator structure inIMRT depends strongly on fields shapes employed for the delivery, potentialcumulative errors on the order of a few per cent can be avoided in fluencecalculations if the proposed method is used.