[Validation of dark band artifact reduction using artifact reduction processing methods with low-dose CT images:...

Area of Science:

• Medical Imaging
• Radiology
• Image Processing

Background:

• Dark band (DB) artifact is a common issue in computed tomography (CT) imaging.
• Accurate CT values are crucial for diagnosis and treatment planning.

Purpose of the Study:

• To validate the effectiveness of iterative reconstruction (IR) in reducing dark band (DB) artifact in abdomen CT.
• To compare IR with other reconstruction methods for DB artifact mitigation.

Main Methods:

• Utilized phantoms (Phantom-A, -B) to simulate abdominal imaging scenarios.
• Evaluated four reconstruction methods: filtered back projection (FBP), adaptive iterative dose reduction using three-dimensional processing (AIDR-3D/MILD), organ-specific reconstruction (OSR), and Boost-3D (Boost).
• Assessed DB artifact using CT values, standard deviation (SD), and profile curves.

Main Results:

• IR methods improved SD values but did not significantly improve CT values for DB artifacts in Phantom-A.
• OSR and Boost methods showed significantly increased CT values for DB artifacts compared to FBP and MILD in Phantom-A.
• No significant changes in CT values were observed due to reconstruction methods in Phantom-B.

Conclusions:

• While IR reduces noise, it can decrease CT values in DB artifacts.
• OSR and Boost methods demonstrate efficacy in improving CT values affected by DB artifacts.
• Reconstruction method choice impacts DB artifact correction in abdomen CT.