Development of Electron Density Phantom for Computed Tomography Image Quality Control Procedures

Abstract

Computed tomography (CT) scanners are imaging technologies that generate high-resolution cross-sectional pictures of anatomical structures with enhanced radiographic contrast. Ensuring high-quality images is crucial for precise diagnostics in medical imaging settings of healthcare institutions by utilising suitable phantoms for quality control protocols. The aim of the research is to develop an electron density phantom for computed tomography image quality protocols. Melt blending and geometric dilution procedures were employed to synthesise distinct tissue-mimicking materials (TMMs) based on physical property (mass density), utilising weight compositions of locally sourced test materials. The mass density of the TMMs was determined. The electron density phantom was designed to be filled with water and accommodate TMM inserts and 3D printed using polylactic acid filament. The phantoms with the inserted TMMs were scanned using GE CT technology at 120 kVp and 200 mA, and the resultant CT images were analysed for image quality evaluation with ImageJ software to identify human tissue based on literature studies. The phantom was subsequently scanned using GE and Philips CT technology following medical imaging protocols for the head and abdomen. It was found that the FTIR analysis of the TMMs had bands that verified the presence of the test materials post-synthesis. The measured mean CT numbers and standard noise of TMMs effectively mimicked several human tissues, including lung of -547.39±14.61 HU, adipose of -73.28±9.60, kidney of 38.71±10.57 HU, muscle of 48.33±11.12 HU, liver of 67.36±9.94 HU, cortical of 1190.16±22.17 HU, and two trabecular bones of 202.05±14.01 HU and 446.75±18.60 HU in comparison to the HU range of human tissues from existing literature works with noise of the TMMs below the 20 HU threshold. The linearity of the CT numbers of TMMs, when scanned using the head and abdomen protocols on the two CT machines, exhibited a strong positive correlation with R2 values of 0.99 and 0.93, respectively,and demonstrated high agreement of the CT numbers at p < 0.05. The results of this research illustrate the precision of the developed electron density phantom, produced through 3D printing with synthesised tissue-mimicking materials, which may be utilised for quality control procedures in medical imaging settings.