High-Throughput Sparsity-Based Inversion Scheme for Optoacoustic Tomography.

Area of Science:

• Medical Imaging
• Computational Imaging
• Signal Processing

Background:

• Sparsity is crucial for data acquisition and image processing, enhancing signal-to-noise and spatio-temporal performance.
• Optoacoustic tomography (OAT) image reconstruction involves computationally intensive inversion of large datasets, especially for volumetric multispectral data.
• High temporal resolution in OAT presents significant computational challenges for accurate image reconstruction.

Purpose of the Study:

• To accelerate accurate model-based optoacoustic inversions.
• To identify and exploit sources of sparsity in forward/inverse models and projection data.
• To enable real-time performance in multi-dimensional optoacoustic imaging.

Main Methods:

• Identifying sparsity in signal, model, and image domains through transformations.
• Applying sparsity-based inversion schemes to experimental optoacoustic data.
• Comparing reconstruction speed and image quality against conventional iterative methods.

Main Results:

• Sparsity-based inversion achieved reconstruction speed enhancements of 40 to 700 times.
• The method preserved image quality comparable to conventional model-based inversions.
• Demonstrated feasibility of significantly faster model-based reconstruction in OAT.

Conclusions:

• Exploiting sparsity offers a powerful approach to accelerate optoacoustic tomography image reconstruction.
• The developed method significantly reduces computational load, paving the way for real-time applications.
• This advancement is critical for multi-dimensional and high-resolution optoacoustic imaging systems.