[Image quality evaluation of new image reconstruction methods applying the iterative reconstruction]

Area of Science:

• Medical Imaging
• Radiology
• Image Processing

Context:

• Computed Tomography (CT) imaging relies on image quality for accurate diagnosis.
• Iterative reconstruction methods offer potential improvements over traditional filtered back projection (FBP).
• Evaluating novel iterative reconstruction algorithms is crucial for advancing CT technology.

Purpose:

• To assess the image quality of the iterative reconstruction in image space (IRIS) algorithm on a 128-slices multi-detector computed tomography (MDCT) system.
• To quantify noise reduction and evaluate its impact on spatial resolution and low-contrast detectability compared to FBP.
• To determine the effectiveness of IRIS in enhancing diagnostic performance.

Summary:

• IRIS demonstrated a 21% reduction in image noise (standard deviation) and lower noise power spectrum (NPS) in middle and high frequencies compared to FBP.
• High-contrast resolution, measured via modulation transfer function (MTF) using a wire phantom, was preserved. However, middle-contrast resolution showed slight degradation with the bar pattern phantom.
• Low-contrast detectability showed no statistically significant difference between IRIS and FBP, despite noise reduction.

Impact:

• IRIS offers effective noise reduction in MDCT imaging.
• The algorithm maintains high-contrast resolution but may slightly compromise middle-contrast resolution.
• Further research is needed to explore potential benefits in specific clinical applications for low-contrast detectability.