SU-E-J-17: Evaluation of Metal Artifact Reduction in MVCTs Using a Model Based Image Reconstruction Method

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computational Phantoms

Background:

• Metal artifacts, particularly dark shading caused by beam hardening, are a significant challenge in megavoltage CT (MVCT) imaging.
• Traditional filtered-back projection (FBP) techniques struggle to adequately correct for these artifacts, leading to image degradation and potential diagnostic inaccuracies.

Purpose of the Study:

• To assess the efficacy of a model-based iterative image reconstruction algorithm in mitigating metal artifacts within MVCT systems.
• To quantitatively and qualitatively compare the performance of this iterative method against the conventional FBP technique.

Main Methods:

• The iterative maximum likelihood polychromatic algorithm for CT (IMPACT) was employed, incorporating pair/triplet production and energy-dependent detector responses.
• Accurate modeling of beam spectra and detector energy-dependent gain was achieved through constrained optimization and direct measurement.
• Scans were performed on a phantom with steel rods and varying density inserts using both bench-top and TomotherapyTM MVCT systems, with FBP serving as a baseline.

Main Results:

• FBP reconstruction exhibited noticeable dark shading between steel rods, causing up to 8% underestimation of electron density with increasing insert density.
• IMPACT reconstruction effectively removed the dark shading artifact, restoring a uniform background and yielding attenuation coefficients close to theoretical values.
• The iterative algorithm demonstrated superior performance in artifact reduction compared to FBP.

Conclusions:

• Iterative reconstruction algorithms like IMPACT can successfully eliminate beam hardening-induced metal artifacts in MVCT imaging.
• Precise modeling of detector response and physical processes is critical for the effective implementation of such iterative algorithms.