IMAGE SEGMENTATION AND VOXELIZATION TECHNIQUES FOR DOSE ASSESSMENT IN X-RAY DIAGNOSTIC PROCEDURES
Computational voxel phantoms are models of the human anatomy used in the field of radiation protection, medical imaging and radiotherapy that enables evaluation of organ doses with a high degree of precision. The gold standard of radiation dosimetry would be to obtain a computational model for each patient involved in radiation processes. Having this in mind, the aim of this work was to try to improve the implementation of a computational voxel phantom starting with a physical one for organ dose assessment and imaging studies in the X-ray diagnostic. The first step was devoted to the segmentation of CT images through thresholding methods and region growing algorithms. The second step consisted in developing a MC model for validation, dose and imaging purposes. Finally, an analysis of SNR for different calcification sizes showed the optimal energy (about 60 keV) that maximizes the image quality for 0.7 cm thick calcification detection.