Development and evaluation of an independent system for absorbed dose calculations in radiotherapy

Sammanfattning: The aim of this work was to develop, implement and evaluate an independent system with which to calculate the absorbed dose, delivered by high-energy X-ray beams, to the prescription point and the depth of dose maximum. The introduction of such a system in the clinical routine may help ensure high-quality treatment and avoidance of errors which may jeopardise the clinical outcome of the treatment (i.e. under- or overdose). A set of equations for calculating the absorbed dose to the prescription point was compiled in a software application (“HandCalc”), which is completely independent of the treatment planning system (TPS). For instance, HandCalc includes models to calculate the absorbed dose from photons scattered in the patient, the transmission of the primary kerma in the patient, the variation of the primary kerma in air with collimator setting (i.e. head scatter), and corrections for heterogeneities in the patient. A new expression for the transmission of the primary kerma in the patient was derived in which the coefficients are strictly defined (and given a physical interpretation) by the first two moments of the spectral distribution of the incident beam. Further investigations also revealed that these moments can be used to determine water-to-air stopping power ratios more accurately than other beam quality indices. In practice, the moments are derived from “in-air equivalent”, narrow-beam measurements using a mini-phantom. The degree of in-air equivalence was investigated with Monte Carlo simulations, which showed that the optimum measurement depth in a mini-phantom is somewhat below the depth of dose maximum. Based upon comparisons with measurements and the TPS, a clinical action level of +/- 4% was chosen for HandCalc. Deviations greater than this are, with all probability, due to erroneous handling of the patient dataset during the preparation phase. An “entrance dose factor” was added in order to correct the dose calculations at the depth of dose maximum where electron equilibrium has not been established. The entrance dose factor was found to vary with beam quality and collimator setting, while no variation was detected with the presence of an acrylic tray (for block support) or with the source-surface distance (SSD). HandCalc was implemented in a hand-held PC which makes dose calculations inside the treatment room at the time of administration of the first fraction possible. An important feature of HandCalc is the built-in report function, which logs results from the calculation for later evaluation. In a study including 700 patients, deviations greater than the action level were found to be due either to limitations in HandCalc or to a systematic deviation between the planned and measured SSD. HandCalc has proven to be a fast and accurate tool for independent dose calculations inside the treatment room and it requires only a limited amount of extra time for the user to perform the calculations. Thus, it can easily be incorporated as part of the daily clinical quality control programme in order to prevent errors which may jeopardise the clinical outcome of the treatment.