Optimization of chest CT protocols based on pixel image matrix, kernels and iterative reconstruction levels - A...

Area of Science:

• Radiology
• Medical Imaging
• Computed Tomography

Background:

• Image reconstruction techniques are vital for optimizing diagnostic accuracy in chest CT.
• Evaluating the impact of matrix size, reconstruction kernels, and iterative reconstruction (IR) levels is essential for improving image quality.

Purpose of the Study:

• To investigate how high matrix image reconstruction, combined with various reconstruction kernels and IR levels, affects image quality in chest CT.
• To determine the optimal parameters for enhanced diagnostic performance in chest CT imaging.

Main Methods:

• Utilized anthropomorphic chest and Catphan® 600 phantoms scanned with a dual-source scanner.
• Performed reconstructions using 768×768 and 1024×1024 matrices with combinations of three kernels and five IR levels.
• Assessed image quality using Signal Difference to Noise Ratio (SdNR) and spatial resolution (lp/cm), with statistical analysis via linear regression and Cohen's kappa for reader agreement.

Main Results:

• Matrix size did not significantly impact SdNR (p=0.595).
• Reconstruction kernel (p=0.014) and IR level (p=0.001) significantly affected SdNR.
• Highest spatial resolution (9 lp/cm) was achieved with specific kernel and IR combinations, with reader scores favoring certain kernels and higher IR levels.

Conclusions:

• Matrix size does not influence chest CT image quality.
• Reconstruction kernel and IR level significantly impact objective and visual image quality.
• Increased IR levels show potential for improving diagnostic image quality in chest CT.