Evaluation of 10 current image reconstruction algorithms for linear array photoacoustic imaging.

Area of Science:

• Biomedical Optics
• Medical Imaging Technology
• Acoustic Imaging

Background:

• Linear array photoacoustic imaging (PA imaging) systems require accurate reconstruction algorithms to visualize absorbers in tissues.
• Existing reconstruction algorithms lack consistent performance evaluation, hindering direct comparison and selection.
• Ten distinct algorithms (DAS, UBP, pDAS, DMAS, MV, EIGMV, SLSC, GSC, TR, and FD) have been developed for PA imaging.

Purpose of the Study:

• To systematically compare the performance of 10 published image reconstruction algorithms for linear array PA imaging.
• To provide a standardized evaluation framework for these algorithms.
• To assist researchers in selecting optimal algorithms for their specific PA imaging applications.

Main Methods:

• Utilized in-vitro phantom data for controlled experimental conditions.
• Evaluated algorithms based on key performance metrics: lateral resolution, computational time, target detectability, and noise sensitivity.
• Performed a systematic comparative analysis of DAS, UBP, pDAS, DMAS, MV, EIGMV, SLSC, GSC, TR, and FD algorithms.

Main Results:

• Quantitative performance data was generated for each of the 10 algorithms across the evaluated metrics.
• Variations in lateral resolution, computational efficiency, target detectability, and noise resilience were observed among the algorithms.
• Specific algorithms demonstrated superior performance in certain aspects, while others showed trade-offs.

Conclusions:

• The comparative analysis provides crucial insights into the strengths and weaknesses of different PA imaging reconstruction algorithms.
• This study offers a foundation for informed algorithm selection in linear array PA imaging.
• Findings will aid researchers in optimizing image quality and efficiency for their specific applications.