Post-processing sets of tilted CT volumes as a method for metal artifact reduction

Area of Science:

• Medical Imaging
• Radiology
• Computational Imaging

Background:

• Metal implants and foreign bodies cause artifacts in computed tomography (CT) scans.
• These artifacts degrade image quality, complicating treatment planning in external beam therapy.

Purpose of the Study:

• To investigate a novel method for reducing metal artifacts in CT reconstructions.
• To improve image quality for quantitative analysis and radiation therapy planning.

Main Methods:

• Acquisition of tilted CT reconstructions from different angles.
• Combining co-registered CT volumes using voxel-wise mean or median calculation.
• Testing the method with calibration phantoms and human skull models with dental restorations.

Main Results:

• Significant reduction in streaking artifacts and improved visibility of details near metal objects.
• Average bias reduced by up to 95%, and noise/errors reduced by up to 94%.
• Method demonstrated comparable performance to established metal artifact reduction (MAR) algorithms with potential for dose reduction.

Conclusions:

• A computationally inexpensive post-processing method for MAR was developed.
• The method requires complementary scans but no a priori information.
• Artifact reduction was achieved without compromising patient dose, offering a viable solution for improved CT imaging.