Pengcheng Long

Development of Chinese Female Computational Phantom Rad-Human and Its Application in Radiation Dosimetry Assessment

Abstract Dose conversion coefficients are important physical quantities in radiation dosimetry assessment and can be derived from Monte Carlo simulation based on human computational phantoms. In order to accurately evaluate the dose to a human body especially for a Chinese female, a precise whole-body Chinese female computational phantom named Rad-Human was constructed based on high-resolution digital color slice images of an adult female body. Rad-Human includes 46 tissues and organs with a minimum voxel size of 0.15 × 0.15 × 0.25 mm for head and neck and 0.15 × 0.15 × 0.5 mm for other regions, and it contains more than 28.8 billion voxels. Conversion coefficients and effective doses of external radiation, specific absorbed fractions, and S values of internal radiation for different energies for Rad-Human were calculated. The calculated dose conversion coefficients were reasonable comparing and analyzing the relationship between dose and organ characteristics with those values of the International Commission on Radiological Protection (ICRP) reference phantom. Based on the information and simulation results of Rad-Human, a set of more complete data of dose conversion coefficients in the radiation field was constructed for a Chinese adult female. Dose discrepancies that were observed were due to differences of body structures between the two phantoms. The differences of dose conversion coefficients between Rad-Human and the ICRP reference phantom demonstrate that Rad-Human can more accurately assess the exposure dose especially for a Chinese female.

Human Organ Geometry Construction from Segmented Images

The human voxel phantom containing large number of voxels is laborious for three dimensional (3D) reconstructions in Monte Carlo simulation applications. A fast and accurate geometry construction method was put forward in present study. The voxels of the phantom was processed as follows: (1) Assigned voxels of one certain organ with 3D coordinates. (2)Based on the coordinate matrix, the voxels were 3D merged into large cuboids. (3)Extract the vertical plane of the surface of the organ in six directions. (4) Planes were used to enclose the organ for the geometry reconstruction. In the process, the voxels was indexed in accordance with the coordinates in order to establish and achieve traversals. With this method, the whole-body phantom of a Chinese radiation virtual human (Rad-HUMAN) was effectively constructed. The geometry described by planes could represent the accurate appearance and precise interior structure of the whole-body phantom. Without ambiguity and self-crossing, it can faithfully describe the voxel pattern of the phantom which can be effectively visualized.

A Method for 3D Structured Data Set Regulation Based on Image

Huge amount of three-dimensional structured data set is generated in neutronics analysis. As for the visualization algorithms are mostly designed for the regular data set, so three-dimensional structured data set needs to be regulated at first. In this paper, in order to regulating three structured data set efficiently, regulation methods were studied, and a regulation method based on dimensional reduction and image technique was presented compared with the traditional regulation method based on interpolation methods. Testing results showed the high effectiveness of the regulation method proposed. The method has been integrated into the scientific visualization program SVIP self-developed by FDS Team

Interactive Visualization of MCNP-Based Neutronics Models

MCNP is a general-purpose Monte Carlo N-Particle code that is widely used in many fields such as radiation protection, dosimetry, radiation shielding, radiography, medical physics, nuclear criticality safety, etc. It can treat an arbitrary three-dimensional configuration of materials in geometric cells. However, it is difficult and time-consuming to visualize and verify the MCNP input file. In order to conveniently check and analyze the MCNP input file, with the development of computer technology, on the basis of mature geometry engine, OOP (Object-Oriented Programming) technique, FDS Team implemented interactive visualization system converting geometry description of MCNP input file into 3D CAD models, which can facilitate to check and analyze the input file. Furthermore, the CAD models generating from the input file can be imported into other commercial CAD software for editing, updating and reusing, it reduces the time of modeling. Now, the interactive visualization system has been integrated into MCAM, as one of its functions, being applied to nuclear analysis.