Abstract
Combining the diffusive laser excitation and the photoacoustic signals detection, photoacoustic computed tomography (PACT) is uniquely suited for deep tissue imaging. A diffraction-limited ultrasound point detector is highly desirable for maximizing the spatial resolution and the field-of-view of the reconstructed volumetric images. Among all the available ultrasound detectors, micro-ring resonator (MRR) based ultrasound detectors offer the lowest area-normalized limit of detection (nLOD) in a miniature form-factor, making it an ideal candidate as an ultrasound point detector. However, despite their wide adoption for photoacoustic imaging, the underlying signal transduction process has not been systematically studied yet. Here we report a comprehensive theoretical model capturing the transduction of incident acoustic signals into digital data, and the associated noise propagation process, using experimentally calibrated key process parameters. The theoretical model quantifies the signal-to-noise ratio (SNR) and the nLOD under the influence of the key process variables, including the quality factor (Q-factor) of the MRR and the driving wavelength. While asserting the need for higher Q-factors, the theoretical model further quantifies the optimal driving wavelength for optimizing the nLOD. Given the MRR with a Q-factor of 1 × 10(5), the theoretical model predicts an optimal SNR of 30.1 dB and a corresponding nLOD of 3.75 × 10(-2) mPa mm(2)/Hz(1/2), which are in good agreement with the experimental measurements of 31.0 dB and 3.39 × 10(-2) mPa mm(2)/Hz(1/2), respectively. The reported theoretical model can be used in guiding the optimization of MRR-based ultrasonic detectors and PA experimental conditions, in attaining higher imaging resolution and contrast. The optimized operating condition has been further validated by performing PACT imaging of a human hair phantom.