A methodology for image quality evaluation of advanced CT systems

Area of Science:

• Medical Imaging Physics
• Radiological Engineering

Background:

• Modern computed tomography (CT) systems utilize advanced techniques like tube current modulation and iterative reconstruction to optimize image quality and radiation dose.
• Quantifying the performance of these techniques across diverse clinical scenarios, including varying body sizes and contrast levels, is crucial for accurate diagnosis.

Purpose of the Study:

• To develop and validate a phantom-based method for comprehensively quantifying the performance of tube current modulation and iterative reconstruction in CT systems.
• To assess key image quality metrics such as resolution, Hounsfield unit (HU) accuracy, noise, and noise texture under varying conditions of contrast, dose, and body size.

Main Methods:

• A polyethylene phantom was designed to simulate four body sizes, featuring uniform sections for noise-power spectrum (NPS) analysis and rod sections for Hounsfield unit (HU) and task-based modulation transfer function (TTF) measurements.
• The phantom was scanned using a clinical CT system with varied tube current modulation settings and reconstruction algorithms (filtered back projection [FBP] and iterative reconstruction [ASIR30]).
• A dedicated image quality analysis program processed phantom data to calculate NPS, TTF, HU, and pixel noise as a function of contrast, dose, and body size.

Main Results:

• The phantom accurately represented different body sizes, with tube current and noise varying as expected with phantom size.
• Iterative reconstruction demonstrated a contrast-dependent TTF, outperforming FBP at low noise levels, though this advantage diminished at higher noise levels.
• Iterative reconstruction resulted in an approximate 30% decrease in NPS magnitude and a shift in peak frequency compared to FBP.

Conclusions:

• A robust phantom and analysis software were successfully developed for evaluating CT image quality across a spectrum of contrasts, doses, and body sizes.
• This platform enables detailed characterization of reconstruction algorithms and tube current modulation techniques through precise NPS, TTF, HU, and noise measurements.