Six iterative reconstruction algorithms in brain CT: a phantom study on image quality at different radiation dose...

Area of Science:

• Radiology
• Medical Imaging
• Computed Tomography

Background:

• Iterative reconstruction (IR) techniques offer potential advancements in CT image quality.
• Evaluating various IR algorithms is crucial for optimizing brain CT protocols.

Purpose of the Study:

• To assess the image quality performance of six distinct IR algorithms across four CT systems.
• To compare IR algorithms against conventional filtered back-projection (FBP) at varying radiation doses and iterative settings.

Main Methods:

• An image quality phantom was scanned on four CT systems at four dose levels.
• Acquisitions were reconstructed using FBP and six IR algorithms (statistical and model-based).
• Image quality parameters included CT numbers, uniformity, noise, noise-power spectra, and contrast/spatial resolution.

Main Results:

• All IR algorithms reduced noise compared to FBP, with greater noise reduction at higher IR levels.
• IR algorithms altered noise distribution, particularly model-based ones.
• Low-contrast resolution improved with all IR algorithms, while spatial resolution was enhanced by model-based and one statistical IR algorithm.

Conclusions:

• Statistical IR algorithms generally improved brain CT image quality over FBP.
• One model-based IR algorithm provided further significant improvements in image quality criteria.