The Research Of Photoacoustic Signal Processing Based On Empirical Mode Decomposition

There is an increasing interest for photoacoutic imaging systems excited by continuous wave?CW?lasers due to the advantages of effective cost and portable potential.However,compared to the powerful pulse laser,the CW lasers provides less light energy,thus resulting in weaker photoacoustic signals.The background noise further interferes the photoacoustic signals in the process of generation,propagation and detection.This will increase the difficulty of the photoacoustic image reconstruction.In order to optimise target photoacoustic imaging results,we focused on the phtoacoustic signal denoising processing based on the empirical mode decomposition.According to the characteristics of the reference signal used in continuous wave laser photoacoustic imaging system,the method of signal reconstruction based on the frequency selection of the intrinsic modal function components and the cross correlation of reference signals are proposed.The simulated data and the measured data of the photoacoustic imaging system demonstrated the denoising effect of the two reconstruction methods respectively.The main contents of this paper are as follows:1.In this paper,we use the frequency selective reconstruction method based on the intrinsic modal function component,utilizing the adaptability of the empirical mode decomposition method,the intuition of each component of the signal after decomposition,and the excitation characteristics of the continuous wave photoacoustic imaging system.Empirical mode decomposition is applied to single component and multi-component signals after adding random noise in simulation experiment.With this reconstruction method,the components obtained by the decomposition are reconstructed.The simulation results show that the signal-to-noise ratio is improved from 4.23 dB to 9.52 dB after multi-component signal reconstruction.2.Since the frequency selection method based on the intrinsic modal function component needs to perform Hilbert spectrum analysis of multiple decomposition signals,this process is time consuming when the signal data volume is large.In order to improve the reconstruction efficiency of the photoacoustic signal,a reconstruction method based on cross-correlation processing of the reference signal is proposed.A group of non stationary signals and linear frequency sweep signals with noise are simulated and analyzed,and the appropriate intrinsic modal function components are selected to denoise the signal by this method.The correlation coefficient and root mean square error are selected to analyze the noise reduction effect of the simulated linear sweep signal before and after reconstruction.The simulation results show that the correlation coefficient of the reconstructed linear sweep signal is improved and the root mean square error is reduced.3.The actual photoacoustic signal is processed and analyzed,and the sample in photoacoustic experiment is carbon rod biomimetic.The photoacoustic signal is first processed by coherent averaging,and then the empirical mode decomposition method is used to decompose the photoacoustic signal of the carbon rod biomimetic.It is verified that the full-component reconstruction error of the photoacoustic signal after empirical mode decomposition is in the range of 10^-17.The reconstruction method based on the intrinsic modal function component frequency selection and the reconstruction method based on the reference signal cross correlation processing are used to filter the noise of the carbon rod biomimetic photoacoustic signal.The measured data demonstrated that the reconstructed components of these two methods are the same,but the signal process based on the reference signal cross-correlation processing reconstruction method is more time-saving.Finally,the correlation coefficient,root mean square error,and signal-to-noise ratio are used as evaluation criteria to analyze the reconstruction results,and the comparison of the reconstructed image contrast of photoacoustic signal before and after reconstruction is given.It is concluded that the signal-to-noise ratio of the photoacoustic signal is improved from 29.71 dB to 37.01 dB after reconstructing by the reconstruction method based on the reference signal cross-correlation processing.