Absorbed dose assessment in the presence of tissue heterogeneities in external radiotherapy

Abstract

The absorbed dose assessment in the presence of tissue heterogeneities in external radiotherapy is an issue that has concerned the medical physics community for almost three decades and it is still a matter of concern. Aiming to obtain dose distributions in clinically-acceptable computation times, analytical dose calculation algorithms integrated in treatment planning systems based their calculations on water-equivalent properties and elemental compositions of each material are disregarded despite the fact that radiation interaction processes strongly depend on them. This approximation provides reasonable accuracy in water-like tissues but the reliability of predicted dose distributions in the patient might be questioned when the radiation beam is traversing complex density heterogeneities, such as air, lung or bone. Experimental verification of dose calculation algorithms is essential and ionization chambers (IC) are the reference detectors for this purpose. However, correction factors to determine the absorbed dose in materials other than water are unknown for most IC types and therefore, they cannot procure reliable measurements in heterogeneous media. Monte Carlo (MC) simulations offer a high precision in dose calculation by tracking all particles individually taking into account the specific properties of each material. Unfortunately, accuracy and computation speed are inversely proportional and MC-based approaches generally entail long calculation times, unaffordable in the clinical routine. Nevertheless, for the cases where the expected errors in the predicted dose distributions during treatment planning are significant, i.e. when the radiation beam path is highly inhomogeneous, the benefit of resorting to MC dose calculations to achieve higher accuracy would be undoubtedly worth a presumably long computation time. In this thesis the suitability of several detectors to accurately determine the absorbed dose in the presence of high-density heterogeneities was evaluated. Ultra-thin thermoluminescent detectors (TLDs) and radiochromic films were considered as potential candidates for entailing low perturbation effects. MC dose calculations enabled to validate and understand the experimental results. Further, both dosimetric techniques were employed to thoroughly examine the behavior of a recently-released non-analytical dose calculation algorithm (AXB)¿which copes with the elemental composition of materials and thus, is claimed to yield promising results¿in heterogeneous phantoms. Finally, a fast algorithm named the heterogeneity index (HI) was developed to quantify the level of patient tissue heterogeneities traversed by the radiotherapy beam. The validity of this HI to easily predict the accuracy of dose distributions based on analytical dose calculations was analyzed by evaluating the correlation between the HI and the dose uncertainties estimated by using MC as the reference. The results show that a detector of 50µm thickness can provide reliable absorbed dose measurements in high-density heterogeneities since perturbation correction factors are unneeded. AXB was found to provide comparable accuracy to MC dose calculations in the presence of heterogeneities but uncertainties in the material assignment procedure might lead to significant changes in the dose distributions, which deserves a word of caution when carrying out experimental verifications. Finally, HI was found to be a fast and good indicator for the accuracy of dose delivery in terms of tumor dose coverage. Accordingly, HI can be implemented in the clinical routine to decide whether or not a MC dose recalculation of the plan should be considered to ensure that dose uncertainties are kept within tolerance levels. In conclusion, this thesis work tackled the main concerns on the absorbed dose calculation and measurement in the presence of tissue heterogeneities.