A new CT based tissue automatic segmentation method for absorbed dose calculation in preclinical radiation therapy
Abstract
Tissue segmentation in CT images is a key parameter for accurate dose calculation in external beam radiation therapy. Tissue segmentation methods used in clinical practice cannot be applied to preclinical practice due to both CT energy (40kVp) and treatment beam energy (100- 300kVp). At MV energy range, tissue segmentation is based on the conversion of HU into densities (HU, ρ). At medium energy range, this conversion method does not allow to achieve a satisfactory accuracy. Taking into account tissue heterogeneities requires to define the tissue elemental composition (EC). We developed a new tissue segmentation method (HU, ρ, EC) allowing the automatic assignment of density and elemental composition from HU. The (HU, ρ, EC) method is dedicated to preclinical practice to reach accurate dose distributions in small animals.