[DR image denoising based on Laplace-Impact mixture model]

Area of Science:

• Digital image processing
• Medical imaging analysis
• Wavelet transforms

Context:

• Diagnostic radiography (DR) imaging is crucial for medical diagnosis.
• Image noise significantly degrades the quality and diagnostic accuracy of DR images.
• Existing denoising algorithms face challenges in effectively removing noise while preserving image details.

Purpose:

• To propose a novel denoising algorithm for DR images.
• To utilize a Laplace-Impact mixture model in the dual-tree complex wavelet domain.
• To develop minimum mean squared error (MMSE) estimators leveraging coefficient correlations.

Summary:

• A new DR image denoising algorithm is presented, employing a Laplace-Impact mixture model within the dual-tree complex wavelet domain.
• The algorithm constructs a probability density function using local variance to accurately model high-frequency subband coefficients.
• It incorporates a novel MMSE estimation method that exploits the correlation between adjacent wavelet coefficients.

Impact:

• The proposed algorithm demonstrates superior performance compared to state-of-the-art methods like Bayes least squared Gaussian scale mixture and Laplace prior.
• This advancement offers improved image quality for diagnostic radiography, potentially enhancing diagnostic accuracy.
• The novel approach provides a more effective solution for noise reduction in medical imaging applications.