Liu Liangbing, Tao Chao, Liu XiaoJun, Deng Mingxi, Wang Senhua, Liu Jun
Opt Express. 2015 Oct 19;23(21):26969-77. doi: 10.1364/OE.23.026969.
Photoacoustic tomography is a promising and rapidly developed methodology of biomedical imaging. It confronts an increasing urgent problem to reconstruct the image from weak and noisy photoacoustic signals, owing to its high benefit in extending the imaging depth and decreasing the dose of laser exposure. Based on the time-domain characteristics of photoacoustic signals, a pulse decomposition algorithm is proposed to reconstruct a photoacoustic image from signals with low signal-to-noise ratio. In this method, a photoacoustic signal is decomposed as the weighted summation of a set of pulses in the time-domain. Images are reconstructed from the weight factors, which are directly related to the optical absorption coefficient. Both simulation and experiment are conducted to test the performance of the method. Numerical simulations show that when the signal-to-noise ratio is -4 dB, the proposed method decreases the reconstruction error to about 17%, in comparison with the conventional back-projection method. Moreover, it can produce acceptable images even when the signal-to-noise ratio is decreased to -10 dB. Experiments show that, when the laser influence level is low, the proposed method achieves a relatively clean image of a hair phantom with some well preserved pattern details. The proposed method demonstrates imaging potential of photoacoustic tomography in expanding applications.
光声层析成像术是一种很有前景且发展迅速的生物医学成像方法。由于它在增加成像深度和降低激光照射剂量方面具有显著优势，从微弱且带噪声的光声信号中重建图像这一问题变得日益紧迫。基于光声信号的时域特性，提出了一种脉冲分解算法，用于从低信噪比的信号中重建光声图像。在该方法中，光声信号在时域中被分解为一组脉冲的加权和。图像由与光吸收系数直接相关的权重因子重建得到。通过仿真和实验对该方法的性能进行了测试。数值模拟表明，当信噪比为 -4 dB 时，与传统的反投影方法相比，该方法将重建误差降低到了约 17%。此外，即使信噪比降至 -10 dB，它也能生成可接受的图像。实验表明，当激光影响水平较低时，该方法能够获得毛发模型相对清晰的图像，且一些图案细节得到了较好的保留。该方法展示了光声层析成像术在拓展应用方面的成像潜力。
