Doctoral Dissertations

Keywords and Phrases

Computed tomography; Deterministic; Discrete ordinate; GPU; Linear Boltzmann equation


"Computed tomography (CT) has become pervasive in medical diagnostics as improved imaging techniques and processing algorithms provide higher quality information to doctors. However, the exponentially increasing usage of CT has raised concerns regarding long term low-dose radiological risks.

Currently, the dose to patients is computed using Monte Carlo methods and experimental tests. In other areas of radiation transport, deterministic codes have been shown to be much faster than Monte Carlo codes.

Currently, no deterministic methodology exists to automatically generate a spatially distributed dose profile from a CT voxel phantom. This work proposes a new code, Discrete Ordinate CT Organ Dose Simulator (DOCTORS) which utilizes a GPU accelerated raytracer and discrete ordinate solver to compute photon flux in the patient. The flux is then converted to dose.

The DOCTORS code was benchmarked against MCNP6 and found to have good qualitative agreement using both a water phantom and a realistic patient phantom. DOCTORS was also found to be much faster than MCNP6; MCNP takes hours to compute flux profiles that take less than a minute using DOCTORS.

A GPU algorithm was implemented that speeds up the DOCTORS code by a factor of up to nearly 40 for large problems. GPU acceleration was found to benefit smaller problems much less. Speedup was seen in both single precision and double precision problems"--Abstract, page iii.


Liu, Xin (Mining & Nuclear Engr)

Committee Member(s)

Alajo, Ayodeji Babatunde
Lee, Hyoung-Koo
Mueller, Gary Edward, 1954-
Erçal, Fikret


Nuclear Engineering and Radiation Science

Degree Name

Ph. D. in Nuclear Engineering


National Research Council


Work is supported by NRC grant NRC-HQ-13-G-38-0026.


Missouri University of Science and Technology

Publication Date

Summer 2017


xii, 103 pages

Note about bibliography

Includes bibliographic references (pages 95-102).


© 2017 Edward T. Norris, All rights reserved.

Document Type

Dissertation - Open Access

File Type




Thesis Number

T 11501

Electronic OCLC #