Evaluation of Eclipse(c) Monte Carlo dose calculation for clinical electron beams using heterogeneous phantoms.

摘要

Accurate dose calculations of photon and electron transport in tissue materials are an important step in the appropriate delivery of cancer radiotherapy. Various commercial treatment planning systems used in radiotherapy provide algorithms for fast dose calculations. It is the responsibility of medical physicists to commission and evaluates these algorithms. In this work, we have evaluated the electron Monte Carlo (MC) algorithm in Eclipse (c) using solid water phantoms with various tissue heterogeneities (water, lung, cortical bone, air) embedded, and using CT data from a real patient. For heterogeneous phantoms, the evaluation is done by comparing dose profiles and percent depth doses (PDDs) calculated on Eclipse with measurements, and with MC simulations using DOSXYZnrc(c). Measurements of dose profiles and PDDs are taken using EBT Gafchromic(c) films, and we have developed a piece of software in Matlab(c) for extracting dose from EBT Gafchromic(c) films. For the real patient case, we use DOSXYZnrc(c) results as a benchmark against which Eclipse is evaluated. Although Eclipse(c) has been evaluated previously, the originality of the present work lays on the use of digitally reproduced phantom copies on Eclipse and DOSXYZnrc (c) instead of CT scanned phantoms, the use of absolute dose for all comparisons, and the consideration of a real clinical patient. In addition, we have developed a tool for extracting absolute dose profiles and PDDs from EBT Gafchromic(c) films. Our results indicate that, MC results agree in general better with measurements (within 5% or less) than Eclipse MC, whose discrepancies with measurements can be as high as 15% for physical phantoms used and as high as 10% in the case of real patient CT data. Largest discrepancies between measurement (or MC) and Eclipse(c) MC occur at depths near and below tissue heterogeneities with relatively sharp density gradients. The slightly better performance of Eclipse(c) for the real patient case is related to the smoother changes in heterogeneities densities in the human body as compared to sharp changes in our physical phantoms.