Assessing the performance and limits of an optoacoustic image reconstruction algorithm using computer-simulated experimental measurements

An optoacoustic imaging algorithm uses backprojection of the time-resolved velocity potential to map the sources of pressure waves detected by an array of acoustic detectors. The relationship between pressure (P) and velocity potential (vp) is P = -rho*d(vp)/dt, or vp = -Integral(P dt)/rho where rho is density, t is time, and d(vp)/dt is the time derivative of the velocity potential. In a forward calculation, a computer simulation can predict the complex pressure waves arriving at each of an array of detectors due to any arbitrary spatial distribution of initial pressure generation. In an inverse calculation, such computer-simulated experimental measurements are used to spatially map the initial pressure source which in optoacoustic imaging corresponds to the initial distribution of pulsed laser energy deposition. Hence, the performance of the inverse calculation as an image reconstruction algorithm could be tested using ideal computer-simulated data. In this report, the image reconstruction algorithm was systematically tested to specify the expected performance under optimal conditions, to illustrate the source/detector geometries that frustrate image reconstruction, and to assess the impact of measurement noise. Typical reconstruction errors were +/- 20%.