A computationally efficient FEM platform for comprehensive simulations of photoacoustic imaging

This study introduces a comprehensive finite element method (FEM) platform to overcome limitations in photoacoustic imaging (PAI) simulations, addressing challenges associated with the simplified numerical methods and rudimentary geometries of existing simulators. The objective is to develop a physics-based numerical simulation method that comprehensively models the entire PAI process, encompassing the various physics processes involved from the initial laser irradiation to the final image reconstruction stage, and producing results that closely replicate real-world scenarios.

The proposed comprehensive simulation platform models the physics of ray optics, bioheat transfer, solid mechanics, elastic waves, and pressure acoustics, encompassing all the various physical processes involved in PAI. This platform employs time-explicit numerical methods, making it computationally efficient and attractive for preclinical analyses. The method was validated by comparing the results of FEM simulations with those from k-wave simulations and experimental tests. The simulations focus on an anatomically realistic breast phantom to demonstrate the induced effects of laser irradiation.

The FEM simulation results revealed that laser irradiation caused a slight temperature increase of approximately 0.6 °C in the tumor area. This temperature increase led to the generation of a maximum pressure stress of 853,000 N m^-2 due to thermoelastic expansion, resulting in the production of acoustic waves with a maximum acoustic pressure of 446 kPa after 2 μs of propagation. These acoustic waves propagate, and are detected by a transducer for subsequent image reconstruction. The reported findings highlight the platform's high precision in simulating PAI, including all of its intermediate steps.

The developed FEM platform is versatile across diverse scenarios, making it a powerful tool for various applications such as PAI simulations of different body parts, evaluation of various beamforming methods, and consideration of different transducer types. The applications of the platform include temperature monitoring during hyperthermia therapy. This simulation method also has significant potential for training machine-learning and deep-learning models.