Efficient correction for CT image artifacts caused by objects extending outside the scan field of view.

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computational Imaging

Background:

• Computed tomography (CT) imaging can produce artifacts when the scanned object exceeds the field of view.
• These artifacts manifest as bright peripheral bands, obscuring crucial anatomical details near the scan boundary.
• Obese or improperly positioned patients are common causes of such truncated projection data.

Purpose of the Study:

• To develop and validate a method for eliminating peripheral bright-band artifacts in CT images.
• To address image quality degradation caused by data discontinuity from out-of-field objects.
• To improve the reconstruction of anatomy at the periphery of the CT scan field of view.

Main Methods:

• An adaptive mathematical extrapolation scheme was employed to mitigate data discontinuity in projection data.
• This extrapolation was performed prior to convolution within a filtered backprojection (FBP) algorithm.
• The method was tested on CT projection data from ten patients with bodies extending beyond the scan field of view.

Main Results:

• Conventional FBP reconstruction resulted in prominent bright-band artifacts obscuring peripheral anatomy.
• The proposed extrapolation-based reconstruction method effectively eliminated these artifacts.
• Reconstructed images showed clear peripheral anatomy with accurate attenuation values, comparable to non-artifactual images.

Conclusions:

• The developed reconstruction method successfully eliminates bright-band artifacts caused by out-of-field anatomy in CT imaging.
• The technique effectively reduces boundary data discontinuity without increasing reconstruction time.
• The algorithm shows potential for generalization to region of interest reconstruction from truncated projections and objects with inhomogeneous density.