Research progress of patient-specific organ doses from CT / 中国辐射卫生

@#The radiation risk caused by CT examination is of great concern. Organ dose is considered to be the most significant technical parameter for quantifying the patient radiation dose and assessing the corresponding risk. At present, the methods to obtain patient organ dose caused by CT examination mainly include physical phantom measurement, direct human body measurement, dose conversion coefficient, Monte Carlo simulation, and dose calculation software. Although different methods have their own characteristics and application, the individualization of organ dose is always the goal of radiation protection and dosimetry research. Patient-specific phantom developed with artificial intelligence and GPU-accelerated Monte Carlo simulation make it possible to calculate the patient-specific organ dose, and the patient-specific organ dose extrapolated by the CT detector signal provides a new solution.

Construction of patient-specific computational twins model based on low dose CT images / 中华放射医学与防护杂志

Objective:In order to make up for the gap in the digital model of twins in the study of fetal radiation dosimetry, this study intends to construct a computational twins phantom based on low-dose CT images.Methods:The low-dose CT images of a pregnant patient were segmented by the combination of threshold automatic segmentation and manual segmentation, which were derived into stereolithography files. The maternal contour, fetal contour and bone structure were constructed using 3D modeling software, and the organ mass was adjusted according to the reference values of ICRP Publication 89 and WHO report after the organs were scaled on a specific scale.Results:The first set of Non-Uniform Rational B-Splines (NURBS) surface phantom of twins in China was established. The external contours, bones, size, position and posture of the phantom were completely consistent with those of the real human body. Each fetus had 25 organs or tissues. After optimization and adjustment, the relative deviation between the twin phantom and the reference value of organ mass was less than 10%.Conclusions:In this study, a personalized computational phantom of twins based on low-dose CT images is constructed, which fills the gap in the twin model. It is helpful in estimating the dose by the fetus from ionizing radiation received in the process of radiation diagnosis of pregnant women with twins.