Evaluation of dose prediction in heterogeneity for a new algorithm on Eclipse Planning System

Yi Rong, University of Wisconsin, Madison, WI

Title: Evaluation of dose prediction in heterogeneity for a new algorithm on Eclipse planning system.

Purpose: To validate the performance of the Analytical Anisotropic Algorithm (AAA) on Eclipse^TM treatment planning system (TPS) in heterogeneity by comparing it with measurements and other two conventional algorithms: Pencil Beam Convolution algorithm on Eclipse and Collapse Cone Convolution algorithm on Pinnacle TPS.

Method and Materials: Simple phantoms were used for dose calculation and analysis. They were homogeneous or heterogeneous phantoms composed of solid water and lung equivalent material with various thickness. Phantoms were CT scanned and exposed by the Linac. Measurements were performed using the ion chamber, silicon diode and films. Dose responses for low and high energy photon beams were investigated by comparing the PDD curves and profiles for different field sizes. Point measurements were compared for various thickness of lung material.

Results: AAA dose prediction fits film measurements very well except an up to ±6% discrepancy in the dose profile across the heterogeneous interface. Point measurements support the AAA calculations. AAA provides accurate dose prediction for low energy photon beam. For the high energy beam and small field size, AAA calculation is up to 8% lower than measurements in lung region.

Conclusion: AAA accounts for attenuation corrections and electron transport, and models the deposited dose in lung regions with greater accuracy. It costs shorter time to finish the calculation. AAA can not accurately model the lateral scattering near beam edges in the heterogeneous region, but it still gives a reasonably close prediction.