Combining dual-tree complex wavelets and multiresolution in iterative CT reconstruction with application to metal...

Area of Science:

• Medical Imaging
• Signal Processing
• Computational Science

Background:

• Metal artifacts (MAR) and noise degrade dental CT image quality.
• Traditional wavelet transforms have limitations in directional analysis for artifact removal.
• Complex dual wavelet transform (DT-CWT) offers enhanced directional analysis for improved MAR.

Purpose of the Study:

• To investigate the benefits of DT-CWT for metal artifact reduction (MAR) in dental CT.
• To evaluate the efficiency of DT-CWT in noise suppression and secondary artifact removal.
• To assess the performance of a multiresolution TV (MRTV) regularized inversion algorithm using DT-CWT.

Main Methods:

• The MRTV approach with DT-CWT was tested on a 2D polychromatic jaw phantom model.
• Simulations included Gaussian and Poisson noise under high noise and sparse measurement conditions.
• Results were compared against single-resolution reconstruction, filtered back-projection (FBP), and Haar wavelet basis.

Main Results:

• DT-CWT filtering effectively removed noise without introducing new artifacts post-inpainting.
• Multiresolution levels resulted in a more robust algorithm than varying regularization strength.
• The DT-CWT approach demonstrated superior performance in noise and artifact suppression.

Conclusions:

• Multiresolution reconstruction with DT-CWT is robust for sparse projection data.
• DT-CWT provides superior metal artifact reduction and noise suppression compared to single-resolution and Haar wavelets.
• The DT-CWT based MRTV algorithm enhances dental CT image quality and diagnostic accuracy.