[Formula: see text] gradient minimization for limited-view photoacoustic tomography

Area of Science:

• Biomedical Imaging
• Image Reconstruction
• Photoacoustic Tomography

Background:

• Photoacoustic tomography (PAT) is an advanced imaging modality offering high resolution and contrast.
• Incomplete data acquisition in PAT, due to geometric constraints or undersampling, leads to artifacts in reconstructed images.
• Addressing artifacts in limited-view PAT is crucial for accurate diagnostic imaging.

Purpose of the Study:

• To develop and validate a novel three-stage reconstruction method for limited-view photoacoustic tomography.
• To enhance image quality by suppressing noise and artifacts.
• To preserve edge information in reconstructed PAT images.

Main Methods:

• Introduction of a [Formula: see text] regularization scheme into the PAT reconstruction process.
• A three-stage approach involving gradient descent for initial solution, projection onto a constraint set, and proximal mapping for refinement.
• Validation through simulation experiments on homogeneous media.

Main Results:

• The proposed method demonstrates superior performance compared to classical reconstruction techniques.
• Effective suppression of noise and artifacts in reconstructed images.
• Preservation of edge details, leading to higher fidelity images.

Conclusions:

• The developed three-stage [Formula: see text] regularization method significantly improves image reconstruction quality in limited-view PAT.
• This technique offers a robust solution for mitigating artifacts inherent in undersampled or geometrically limited photoacoustic data.
• The findings support the clinical potential of enhanced PAT imaging for various diagnostic applications.