Enhancing sparse-view photoacoustic tomography with combined virtually parallel projecting and spatially adaptive...

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Computational Imaging

Background:

• Photoacoustic tomography (PAT) requires rapid measurements and robust reconstructions for clinical and preclinical use.
• Sparse-view measurements accelerate data acquisition but pose reconstruction challenges.
• Effective measurement strategies and appropriate reconstruction algorithms are crucial for sparse-view PAT.

Purpose of the Study:

• To develop an iterative sparse-view PAT reconstruction scheme.
• To improve image fidelity and performance with limited measurement data.
• To address the challenges of reconstructing PAT images from sparse sampling.

Main Methods:

• Introduced a virtual parallel-projection concept to match measurement conditions.
• Utilized compressive sensing principles within the reconstruction process.
• Employed non-local spatially adaptive filtering to leverage image similarities for improved recovery.

Main Results:

• The proposed iterative scheme significantly improved reconstructed image quality compared to the universal back-projection method.
• Demonstrated enhanced image fidelity even with a reduced number of measurement positions.
• Validated the approach through simulations and ex vivo experiments.

Conclusions:

• The developed iterative sparse-view PAT reconstruction scheme effectively enhances image quality and fidelity.
• The method shows promise for accelerating PAT data acquisition without compromising diagnostic accuracy.
• This approach is suitable for applications requiring high-quality imaging from limited measurements.