First report of the quality assurance of a dual-energy computed tomography (DECT) system: a phantom study

• Main Author: Low, Reece Cin
• Other Authors: -
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Science::Physics::Radiation physics
• Online Access: https://hdl.handle.net/10356/166248
• Institution: Nanyang Technological University

This report conducts a comprehensive phantom study on the dual-energy computed tomography (DECT) system of General Electric (GE) Revolution™ CT (GE Healthcare, Chicago, Illinois, USA), which will be used by National Cancer Centre Singapore (NCCS). This study is done with the use of a new commercial multi-energy computed tomography (MECT) phantom - Multi-Energy CT Phantom (Sun Nuclear, Model 1472, Melbourne, Florida, USA). Quality assurance (QA) of the DECT system is necessary to ensure that all images acquired are functioning optimally and can be trusted and used with confidence during treatment planning. However, there is currently no published guideline on how to best QA a DECT system. Hence, to QA the DECT system in this study, the functionality of two applications of DECT - monoenergetic reconstruction and ρ - Zeff decomposition - was evaluated. Results show that the HU and effective atomic number measurements obtained using the system were generally accurate. Findings also indicate that there is a dependence of HU accuracy on beam energy, concentration of iodine and calcium, as well as phantom size, and there is also a dependence of effective atomic number accuracy on fitting parameters used. However, since all the HU results obtained were within tolerance and there is a strong correlation between measured and calculated effective atomic numbers, the results obtained are considered acceptable and could be used in clinical practice. This validates the proper functioning of monoenergetic reconstruction and ρ - Zeff decomposition in the DECT system and signifies that the methods and measurement procedures employed in this study can be used in future routine QA procedures.