Spatially-variant image deconvolution for photoacoustic tomography.

Area of Science:

• Biomedical Imaging
• Optical Imaging
• Image Reconstruction

Background:

• Photoacoustic tomography (PAT) systems offer high-resolution, high-contrast biological tissue imaging.
• PAT images often suffer from spatially variant blur and streak artifacts, hindering accurate analysis.
• Current reconstruction algorithms and non-ideal imaging conditions contribute to image degradation.

Purpose of the Study:

• To develop a progressive, two-phase restoration method for improving PAT image quality.
• To address and mitigate spatially variant blur and streak artifacts in PAT images.
• To enhance the diagnostic value of PAT by improving image fidelity.

Main Methods:

• Phase 1: Precisely measured point spread function (PSF) samples were modeled using principal component analysis and radial basis function interpolation to capture spatially variant blur.
• Phase 1: A sparse logarithmic gradient regularized Richardson-Lucy (SLG-RL) algorithm was developed for deblurring.
• Phase 2: A novel deringing method, also based on SLG-RL, was introduced to remove streak artifacts.

Main Results:

• The proposed method successfully modeled the spatially variant PSF.
• Deblurring using the SLG-RL algorithm effectively reduced image blur.
• The deringing method significantly removed streak artifacts from PAT images.
• Evaluations using simulations, phantoms, and in vivo experiments demonstrated substantial image quality improvement.

Conclusions:

• The developed two-phase restoration method offers a significant advancement in PAT image processing.
• This technique effectively tackles common artifacts, leading to clearer and more reliable PAT images.
• The findings suggest broader applicability of this method for enhancing biomedical imaging quality.