Signal processing using wavelet transform in photoacoustic tomography

In order to improve the imaging contrast and resolution in pholoacoustic tomography(PAT), the deconvolution between the transducer impulse response and the recorded photoacoustic(PA) signal of the tissue phantom is often used. The suppression of noise is critical in the deconvolution. Compared with the traditional band-pass filter in Fourier domain, wiener filter is more appropriate for the wide band PA signal. The scaling parameter in wiener filter is hard to determine using the traditional Fourier domain method. To solve the problem, the deconvolution algorithm with wiener filter based on the wavelet transform is presented. The scaling parameter is estimated using discrete wavelet transform(DWT) by its multi-resolution analysis(MRA) ability. The white noise had been effectively suppressed. Both numerical simulation and experimental results demonstrated that the contrast and resolution of PA images had been improved.