A virtual imaging framework for three-dimensional quantitative optoacoustic tomography using stochastic numerical...

Area of Science:

• Biomedical Imaging
• Medical Physics
• Computational Imaging

Background:

• Optoacoustic tomography (OAT) shows promise for breast cancer diagnosis by visualizing tumor angiogenesis and hypoxia.
• Accurate image reconstruction in quantitative OAT (qOAT) is crucial for clinical applications and relies on measurement quality and imager design.
• Virtual imaging provides a cost-effective approach for evaluating system designs and developing computational methods.

Purpose of the Study:

• To present a comprehensive virtual imaging framework for breast qOAT.
• To extend a numerical breast phantom generator with variations in skin tone and inclusion of benign/malignant lesions.
• To enable end-to-end simulation for system design evaluation and computational method development.

Main Methods:

• Developed a virtual imaging framework for breast qOAT.
• Extended a stochastic numerical breast phantom (NBP) generator to include diverse skin tones and pathological lesions.
• Modeled transducer spatial and electro-acoustic impulse responses for end-to-end simulation.
• Made 1020 NBPs and measurement data publicly available.

Main Results:

• Demonstrated the framework's utility through a comparative case study of two system designs.
• Facilitated the creation of a diverse dataset of NBPs and simulated measurements.
• Provided a platform for advancing computational methods and optimizing system design.

Conclusions:

• The developed virtual imaging framework is a versatile tool for breast qOAT research.
• The publicly available dataset accelerates research in optoacoustic and optical imaging.
• The framework supports the optimization of imaging system designs and computational techniques for breast cancer diagnosis.