Image-based iterative compensation of motion artifacts in computed tomography.

Area of Science:

• Medical Imaging
• Computed Tomography (CT)
• Image Reconstruction

Background:

• Patient motion during computed tomography (CT) scans causes significant artifacts, leading to reduced image quality and diagnostic accuracy.
• Free-breathing acquisitions are particularly susceptible to motion-induced artifacts like streaks, structural doubling, and blurring, limiting resolution.

Purpose of the Study:

• To present an iterative method for compensating motion artifacts in slowly rotating CT systems.
• To improve image quality by reducing blurring and enhancing the visualization of anatomical structures affected by patient motion.

Main Methods:

• An iterative approach is proposed to correct motion artifacts.
• The method involves decomposing motion in projection space, reconstructing artifacts in image space, and subtracting them from an initial image.
• The initial image, though potentially blurred, serves as a reference for artifact estimation.

Main Results:

• Experiments using simulated projections demonstrated significant improvements in image quality.
• The diaphragm's border became sharper, and contrast for small lung structures improved.
• Qualitative and quantitative analyses confirmed the effectiveness of the proposed method.

Conclusions:

• The technique reconstructs motion artifacts in image space, avoiding explicit motion estimation.
• This novel approach offers potential for generating sharp, static images on interventional C-arm systems and slow X-ray equipment used in radiotherapy.