Deconvolution based photoacoustic reconstruction with sparsity regularization.

Area of Science:

• Biomedical Imaging
• Photoacoustic Tomography
• Signal Processing

Background:

• Standard photoacoustic tomography (PAT) reconstruction assumes ideal transducers and full surrounding coverage, which are often unmet in practice.
• Limitations in transducer directivity and viewing angles can degrade image reconstruction quality.

Purpose of the Study:

• To develop a novel deconvolution-based photoacoustic reconstruction technique with sparsity regularization (DPARS).
• To address practical limitations of transducer directivity and limited viewing angles in PAT.
• To improve the accuracy and efficiency of photoacoustic image reconstruction.

Main Methods:

• Developed the DPARS algorithm, a semi-analytical approach using deconvolution and sparsity regularization.
• Incorporated transducer directivity effects into projections computed over limited viewing angles.
• Utilized discrete cosine transform for sparse representation of absorber coefficients to enhance numerical stability and reduce computation time.

Main Results:

• DPARS demonstrated improved performance in simulations and experimental 2D/3D data using conventional ultrasound transducers.
• Evaluated using contrast-to-noise ratio and root-mean-square errors, DPARS outperformed the conventional delay-and-sum (DAS) reconstruction method.
• Achieved a one-order-of-magnitude reduction in computation time compared to Tikhonov regularization.

Conclusions:

• DPARS effectively overcomes limitations of traditional PAT reconstruction methods.
• The technique offers superior image reconstruction quality and computational efficiency for photoacoustic imaging.
• DPARS shows significant potential for advancing practical applications of photoacoustic tomography.