Usage of low energy alternatives, such as light emitting diode (LED) and pulsed laser diode (PLD), to solid state lasers as a source of illumination in photoacoustic tomography (PAT) is being actively investigated for several preclinical and clinical applications due to its inherent advantages such as portability, low cost and high pulse repetition frequency (PRF). However, the photoacoustic signal acquired from such low energy sources particularly LED, have weak light absorption-to-ultrasound conversion efficiency. Due to the weak nature of the generated PA signals, noises from external sources, such as, electronic noise, jitter, etc. have a more pronounced detrimental effect on the quality of the reconstructed LED-PAT images resulting in low Signal-to-Ratio (SNR) values. The high PRF of LED-PAT systems is usually exploited to improve the SNR by performing averaging of several hundreds of frames. However, this is a time-consuming procedure and decreases the frame rate achievable using the imaging system. In this work, we demonstrate that by employing signal recovery strategy at sub-λ array locations along with a denoising filter, it is feasible to improve image quality and reduce noise from LED-PAT images acquired using commercially available LED-PAUS imaging system (AcousticX, Cyberdyne Inc., Ibaraki, Japan). This method was tested on phantom and in-vivo data acquired using the above-mentioned system and was found to yield reduction in background noise of up to 12 dB compared to mere averaging Also, by applying the proposed method, we achieved two-fold improvement in lateral resolution.
|