Using simulated data in computationally intelligent photoacoustics

Abstract

A development of simulated data and discussion on potential use in processing photoacoustic measurement data, as well as in solving of an inverse problem are presented in the paper. Simulated experimental values are obtained based on theoretical-mathematical model of photoacoustic response, described in the paper. Credibility of the data is acquired due to the experimental measurements. Simulated data validity was proven with the satisfactory accuracy, precision, and reliability of the several machine learning models which were trained on the designed database and are applied in processing photoacoustic measurement data. Presented results show that if a huge dataset of highquality data is generated, solving issues of inverse photoacoustic problem is successful, computing is cost effective, and process rate of machine learning models is suitable for real time work.