Multi-bandwidth image reconstruction in photoacoustic tomography

Abstract

Photoacoustic tomography (PAT), also referred to optoacoustic tomography, is a hybrid imaging technique that combines an optical contrast mechanism and ultrasonic detection principles. The laser-induced photoacoustic signals in PAT are broadband in nature, but only a bandpass approximation of the signal is recorded by use of a conventional ultrasonic transducers due to its limited bandwidth. To circumvent this, a PAT system has been developed that records photoacoustic signals by use of multiple ultrasonic transducers that possess different central frequencies. In this work, we investigate a sensor fusion methodology for combining the multiple measurements to obtain an estimate of the true photoacoustic signal that is superior to that obtainable by use of any single transducer measurement. From the estimated photoacoustic signals, three-dimensional images of the optical absorption distribution are reconstructed and are found to possess improved accuracy and statistical properties as compared to the single transducer case. Preliminary computer-simulation studies are presented to demonstrate and investigate the proposed method.