Abstract. An algorithm for dynamic multileaf-collimator (dMLC) tracking of a 8 target performing a known a priori, rigid-body motion during volumetric modulated 9 arc therapy (VMAT), has been experimentally validated and applied to investigate

A model has been developed to simulate volumetric modulated arc therapy (VMAT) delivery for Elekta control systems. The model was experimentally validated for static-tumour VMAT delivery and has been applied to the investigation of motion compensation with dynamic multileaf collimator (dMLC) delivery tracking for a series of VMAT lung treatment plans at various control point spacings for five patients. The relative increase in treatment time with dMLC tracking was calculated for four 1D rigid-body motion trajectories, and the effect of the control point spacing, the MLC leaf speed and an increased number of dose levels on the dMLC tracking delivery time evaluated. It has been observed that a faster leaf speed is advantageous for motion trajectories with shorter time periods and larger amplitudes. The accuracy of dMLC tracking was found to increase with a decreased control point spacing and is dependent on the amplitude and time period of the motion trajectory of the target. dMLC tracking is shown to be a promising emerging technology which can confer advantage over breath-hold motion-compensation techniques which more drastically reduce the efficiency of VMAT and are more invasive for the patient.

Abstract. The AccuLeaf mMLC featuring four multileaf-collimator (MLC) banks has been used for the first time for an experimental comparison of conventional twobank with novel four-bank dynamic MLC tracking of a two-dimensional sinusoidal respiratory motion. This comparison was performed for a square aperture, and for three conformal treatment apertures from clinical radiotherapy lung cancer patients. The system latency of this prototype tracking system was evaluated and found to be 1.0 s and the frequency at which MLC positions could be updated, 1 Hz, and therefore accurate MLC tracking of irregular patient motion would be difficult with the system in its current form. The MLC leaf velocity required for two-bank-MLC and fourbank-MLC tracking was evaluated for the apertures studied and a substantial decrease was found in the maximum MLC velocity required when four-banks were used for tracking rather than two. A dosimetric comparison of the two techniques was also performed and minimal difference was found between two-bank-MLC and four-bank-MLC tracking. The use of four MLC banks for dynamic MLC tracking is shown to be potentially advantageous for increasing the delivery efficiency compared with twobank-MLC tracking where difficulties are encountered if large leaf shifts are required to track motion perpendicular to the direction of leaf travel.