A Kalman filtering framework for virtual sensor-enhanced photoacoustic imaging

Area of Science:

• Biomedical optics
• Medical imaging
• Acoustics

Background:

• Photoacoustic imaging (PAI) offers high contrast and resolution but suffers from incomplete angular sampling and noise.
• Existing PAI systems face limitations due to detector array configurations and inherent measurement uncertainties.

Purpose of the Study:

• To develop a model-based Kalman filtering framework for generating virtual sensor measurements in PAI.
• To enhance angular information and improve image reconstruction quality without altering physical hardware.

Main Methods:

• Implemented a Kalman filtering approach to estimate virtual sensor data at intermediate angular positions.
• Exploited acoustic wave propagation and signal temporal coherence for noise-aware, minimum-variance pressure field estimation.
• Validated the method using k-Wave simulations with realistic parameters like finite-aperture detectors and acoustic attenuation.

Main Results:

• The virtual sensing strategy significantly improved structural preservation in PAI reconstructions.
• Quantitative image quality was substantially enhanced compared to traditional interpolation methods.
• Demonstrated improved performance in heterogeneous media and with acoustic attenuation.

Conclusions:

• Kalman-domain virtual sensing is a practical and physically grounded method for augmenting PAI acquisition.
• This approach enhances PAI reconstruction quality without requiring modifications to detector hardware.
• The technique offers a viable solution for overcoming limitations of incomplete angular sampling and measurement noise in PAI.