A flexible system model for real-world OMUS-based photoacoustics

Photoacoustic spectral imaging (PASI) is a promising non-invasive sensing modality revealing molecular contrast in tissue by pulsed optical excitation and ultrasound detection. Despite its potential, PASI still encounters significant challenges due to the demanding requirements on the ultrasound detectors to achieve good image quality at depth in tissue. On a component level, the 100-fold higher sensitivity of imec’s opto-mechanical ultrasound sensor (OMUS) shows great potential to replace conventional sensors for PASI. However, identifying for which (pre-) clinical applications its exquisite component specs can make the difference at system level is very challenging, costly and time consuming. Therefore, we have built a digital twin for PASI systems, including the detailed characteristics of the OMUS, a large data set of tissue properties and the ability to model complex geometries. In this work, we focus on the high-level architecture of the system model and on the experimental validation by showing how well the simulated time traces match in 3 controlled experiments with increasing complexity and relevance. Additionally, we illustrate the resulting good match in the reconstructed images for one case. This flexible and accurate model enables system level design optimization and performance prediction for (pre-)clinical applications.