Evaluación de la cobertura dosimétrica tumoral en tratamientos radioterápicos de cáncer de pulmón

Abstract

The use of radiotherapy techniques based on beam modulation suposes an increment of dose divergences between treatment planning system and more sofisticated calculation algorithm, like Monte Carlo method, specially in situations of high hetereogeneity of tissues. These absorbed dose variations have a special importance on dose distribution at the tumoral volume, creating the possible presence of underdosed regions which implies lost in tumor control probability. Our main task in this thesis project is to know the origin of these dosimetric divergences between calculation systems, and go in deep with the clinical implications in lung cancer clinical cases. For this purpose, we are going to use geant4 as Monte Carlo calculation reference. Keeping the achievement of the aim as main guideline, this work is divided in three parts: first of all we are going to analyse the most general aspects of lung cancer as introduction, including the most frequent radiotherapy techniques for lung cancer treatment. Following we describe the items involved in patient irradiation process, i. e. the lineal accelerator and the treatment planning system. Moreover, we describe the radiochromic film as the detector to employ in dose measurements. Finally, we present the more relevant properties of the Monte Carlo method and geant4 code. On the second part we evaluate sereval geant4 applications developed towards dosimetric validation of the geometrical model of the items which compose the treatment head of the Siemens Oncor Impression Plus linac, installed at the Hospital Universitario Virgen Macarena (Seville, Spain), through the comparison between experimental measurements acquired with ionization chamber in water. Furthermore, we have taken into account the differences between dose distributions calculated with geant4 and the treatment planning system, making comparisons with experimental dose distribution measured with radiochromic film inserted inside a anthropomorfic phantom. This phantom includes several materials with known chemical compositions and physical densities. Hence, these studies defined above represent the academic section of our thesis project. Finally, on the third and last part we define the realistic component of the thesis project. We use DICOM images for developing the geometric models of the phantom and patient through our geant4 tools. In this way we achieve an equality in geometric terms between treatment planning system and Monte Carlo method. First of all, we have analysed the dosimetric differences for the calculations (compared with the experimental values) at the voxelized geometry of the phantom. Lastly we have done the same process for real lung cancer treatments. In this case we have evaluated too the radiobiological consecuences due to dosimetric divergences thanks to the tumor control probability calculation.