Research On Noise Reduction Method Of The In-vivo Photoacoustic Detecting Ring System

Malignant melanoma is a kind of malignant tumor developed from melanocytes with high metastasis rate and high fatality rate.It has developed rapidly in China in recent years.Circulating tumor cells(CTC)is an important marker of tumor metastasis,so the quantitative detection of melanoma circulating cells is of great significance to the circulation and monitoring of cancer.In-vivo photoacoustic detection ring system,based on the In-vivo photoacoustic flow cytometer(PAFC),a number of clinical improvements have been made to achieve the miniaturization of the system,which has strong practicality and good clinical prospects.The in-vivo photoacoustic detection ring system is composed of a photoacoustic signal excitation module,a photoacoustic signal acquisition module,and a laser frequency-doubling navigation module.The high repetition frequency pulse laser with a wavelength of 1064 nm passes through the lens group consisting of a double meniscus lens and a double cemented lens,and is parallelly incident on the objective lens to form a light spot smaller than 1mm2.An ultrasonic transducer with a center frequency of 10 MHz collects the photoacoustic signal emitted by the target cells,and after being amplified and filtered by an amplifier,it is sampled and stored by a Ga Ge acquisition card with a sampling rate of 50 MHz The laser signal positioning and navigation module includes a double meniscus lens,a double cemented lens and a KTP frequency-doubling crystal,which locates the blood vessel through the difference of the photoacoustic signal between the blood vessel and the skin under the532 nm wavelength laser.The photoacoustic signal contains the key information of many cells,and the signal processing and analysis is the key link of the experiment.The original photoacoustic signal is affected by a series of noises such as breathing noise,muscle motion noise,and background noise.Before analyzing the signal,noise reduction must be performed.The commonly used signal noise reduction method is the time-domain average noise reduction method.However,this method relies on multiple measurements of multiple single signals,which is difficult to achieve in clinical experiments and cannot be used when the noise power is greater than or equal to the signal power.Therefore,we refer to the relevant literature and use the wavelet noise reduction method and the empirical mode decomposition(EMD)threshold noise reduction method to perform noise reduction on the signal.On this basis,an improvement to the EMD threshold noise reduction method is proposed.This method implements adaptive decomposition of the photoacoustic signal through EMD,and then divides the eigenmode function(IMF)into discrete Fréchet distance as a criterion.For high-frequency group and low-frequency group,a fixed threshold for noise reduction is adopted for the high-frequency group IMF,and the low-frequency group IMF is subjected to Savitzky-Golay(SG)filtering to reduce noise,and then reconstructed to obtain a noise-reduced signal.The above three noise reduction methods are applied to the noise reduction of monopulse signals.The wavelet threshold noise reduction method and the EMD threshold combined with the SG filter noise reduction method have a very significant effect on improving the signal to noise ratio.The EMD threshold method and the EMD threshold are combined The SG filtering method is stronger for suppressing pure noise.The above three noise reduction methods were used for the data collected in animal experiments.All three methods effectively reduced the false positive rate.The EMD threshold combined with the SG filter noise reduction method reduced the false positive rate from 49.1% to 11.4%.Experiments show that EMD threshold combined with SG filter noise reduction method has achieved good data processing effect.The in-vivo photoacoustic detection ring system provides a new tool for non-invasive,label-free detection of melanoma circulating cells,and plays an important role in the early diagnosis and real-time monitoring of melanoma.In addition,the technology has potential application value in evaluating treatment effects,predicting tumor metastasis,and providing individualized treatment.