Computationally efficient error estimate for evaluation of regularization in photoacoustic tomography

Area of Science:

• Biomedical Imaging
• Computational Imaging
• Medical Physics

Background:

• Model-based image reconstruction in photoacoustic (PA) tomography necessitates explicit regularization.
• Determining optimal regularization parameters is crucial for accurate PA image reconstruction.

Purpose of the Study:

• To develop and validate an error estimate (χ²) minimization-based approach for determining regularization parameters in PA imaging.
• To assess the computational efficiency and quantitative accuracy of the proposed method compared to existing techniques.

Main Methods:

• Proposed an error estimate (χ²) minimization approach for regularization parameter determination.
• Integrated the method within the Lanczos bidiagonalization framework for dimensionality reduction.
• Evaluated the method's performance against state-of-the-art techniques and applied it to Tikhonov, exponential, and nonsmooth regularization.

Main Results:

• The proposed method demonstrated computational speed advantages over current state-of-the-art techniques.
• Achieved comparable quantitative accuracy in reconstructed photoacoustic images.
• Successfully utilized the error estimate (χ²) for parameter determination across diverse regularization methods.

Conclusions:

• The developed error estimate (χ²) minimization approach offers an efficient and accurate solution for regularization parameter selection in PA tomography.
• This method enhances the robustness and applicability of various regularization techniques for improved photoacoustic imaging.