Analyzing For Photoacoustic Signals Detected By In Vivo Photoacoustic Flow Cytometry

In this master project,student improved the analysis for photoacoustic signals detected by in vivo photoacoustic flow cytometry.When biological tissues are irradiated with laser,target cells in tissues would absorb energy,causing temperature increases and volume expansion.During this process,the target cells would emit ultrasonic waves outward.This ultrasonic wave signals were detected by the ultrasonic probe.Based on the photoacoustic effect,we have developed an in vivo photoacoustic flow cytometry(PAFC).The instrument can detect photoacoustic signals of target cells in live animals.In the process of tumor metastatic formation,circulating tumor cells appears in the blood or lymph circulation.Therefore,early detection of circulating tumor cells has great significance for timely treatment of the cancer to reduce cancer death.PAFC can detect circulating melanoma cells and obtain photoacoustic signals from cells effectively.By analyzing and processing the photoacoustic signals,we can obtain the signal characteristic of tumor cells and determine the source of the signal.It is helpful to achieve our goal of early detection of cancer cells.Firstly,we used denoising methods to distinguish melanoma cells signals from background noise.The method including Wavelet denoising and Average denoising.Wavelet denoising can effectively filter the signal of noises;Average denoising can improve signal to noise ratio.Processing of photoacoustic signals could be divided into two parts: the time domain processing and the frequency domain processing.In the time domain,the methods could be divided into threshold evaluation and related parameters statistics.In threshold evaluation,we set the threshold using the formula method.The statistical parameters were FWHM and the gap between peak crest and trough.In the frequency domain,the methods to analyze signals we used include fast Fourier transform(FFT)and filter;calculate power spectrum;analyze of the spectrum shape;calculate the envelope curve and tailing factor;calculate and compile statistics of FWHM in signal spectrum.In the process of calculation,the FFT transform data points can be adjustable;the purpose of filtering is to find specific frequency bands of different signals.When calculating power spectrum,we used different window functions.In analyzing of the spectrum shape,we focus on the position of the main peak in the spectrum and the spacing between a series of peaks.We used the Hilbert transform method and linear interpolation to calculate envelope curve and the tailing factor.With the tailing factor,we sought the differences between different signals.Finally,we calculated the signal peak FWHM in spectrum.We used this parameter to find the difference between the photoacostic signals from different target cells.After the calculation and analysis above,we found a way to judge whether the photoacoustic signals come from melanoma cells.In time domain processing,the "triple standard deviation" threshold method is used to determine whether there are peak signals from melanoma cells.In the frequency domain,we have found that the FWHM values of melanoma signals are concentrated in a certain frequency range.So we calculate the FWHM values in spectrogram of signal peak to compare with the certain range.Combined with these two methods,we can determine that the detected signals were derived from melanoma cells in the circulation system.We can achieve our goal of early detection of melanoma cells.