Attributes of a parametric semi-analytical voxel phantom for computational dosimetry via MCNP-5 code

Radiation therapy (RT) represents a technique for cancer treatment, which can also brings deleterious irreversible effects. In order to minimize this effects it is necessary that the radiation planning improves to a personal level such a way that radiation dose be optimized in the patient. Computational radiation dosimetry using voxel model can bring reality close to the anatomy and morphology of the irradiation subject. Despite of the computational cost, the idea of running personal voxel's model for predicting absorbed dose is attractive. However, to reduce computer time, restrictions in this technique should be included, specially restricting the interest region to be discreted and including it in a semi-analytical model. Thus, the tumor region can be discreted in a voxel model while a semi-analytical simplified phantom reproduces all the human body with its on metric. The paper presents a mathematical model that follows theses principles. A semi-analytical mathematical model is created with parameters provided by the metric of the patient. Such parametric phantom with the anatomic values of the patient can reproduce the external geometry of the patient. It was tested on pediatric oncology. A parametric model of a six years child was developed, in which a head and neck voxel model was prepared and adjusted to the semi-analytical phantom. Such set of data was applied to predicted absorbed dose in radiation protocols in brain. Results of the dose distribution and the features of the phantom were presented. (author)