Statistical iterative reconstruction for streak artefact reduction when using multidetector CT to image the...

Area of Science:

• Medical Imaging
• Computerized Tomography (CT)
• Image Processing

Background:

• Metal artifacts, specifically streak artifacts, are unavoidable in CT images when metallic prosthetic appliances or dental fillings are present in the oral cavity.
• These artifacts can significantly degrade image quality and hinder accurate diagnosis.

Purpose of the Study:

• To develop and evaluate a statistical reconstruction method for reducing metal-induced streak artifacts in multidetector row CT images.
• To improve the diagnostic quality of CT images in the presence of metallic dental materials.

Main Methods:

• A sequential iterative restoration process was employed, utilizing projection data from adjacent, artifact-free CT slices.
• The study examined the Maximum Likelihood-Expectation Maximization (ML-EM) and Ordered Subset-Expectation Maximization (OSEM) algorithms.
• A small region of interest (ROI) approach and the use of a general-purpose graphics processing unit (GPGPU) were also investigated for optimization.

Main Results:

• The sequential processing method effectively reduced metal-induced streak artifacts in multidetector row CT images.
• The OSEM algorithm and the small ROI approach demonstrated reduced processing times without compromising artifact reduction efficacy.
• Implementation on a GPGPU significantly enhanced the processing performance, enabling high-speed artifact correction.

Conclusions:

• Statistical reconstruction offers an effective approach for streak artifact reduction in CT imaging.
• Alternative algorithms like OSEM, combined with techniques such as small ROI selection and GPGPU acceleration, provide efficient and fast artifact correction.
• These advancements improve CT image quality in patients with metallic dental restorations.