Research On Signal-to-noise Ratio Enhancement Method For Fluorescence Endoscopy

Fluorescence Molecular Imaging(FMI)is an effective technology for accurately localizing tumors.The technology can help doctors accurately identify healthy tissues and tumors by using fluorescence probes to label tumors and display fluorescence images of the tumor area in real time.The imaging principle of fluorescence endoscope is the same as that FMI technology.Fluorescent probes is used to label tumor tissues in fluorescence endoscope,which can help doctors to detect tiny lesions,accurately locate tumors,label tumor boundaries,and precisely remove tumors.However,due to the weak fluorescence signal,the fluorescence image quality is poor.In order to improve the image quality of fluorescent endoscope,this paper uses light source modulation technology,and digital lock-in algorithm to simultaneously collect white light images and fluorescent images.At the same time,the image signalto-noise ratio is enhanced and the image quality is improved.Firstly,the light source modulation method and digital lock-in algorithm are developed,and the signal-to-noise ratio enhancement algorithm of fluorescent endoscope image based on fast digital lock-in algorithm is studied.Finally,and the signal-to-noise of the white images and the fluorescence images,which are processed by lock-in algorithm and not dealt with lock-in algorithm,are computed,compared and analyzed.In the algorithm research experiment,when 20 dB noise is added to original image,the signal-to-noise ratios of white images and fluorescence images with digital lock-in algorithm processing are 23.98 dB and 22.10 dB,respectively.And the signalto-noise ratios of white images and fluorescence images without digital lock-in algorithm processing are 16.99 dB and 15.11 dB,respectively.Compared with the image signal-to-noise ratio without lock-in image algorithm processing,the signal-tonoise ratio of the white images and the fluorescence images with digital lock-in algorithm processing are increased by 41.1% and 46.3%,respectively.Secondly,in order to verify the effect of light source modulation technology and digital lock-in algorithm on image signal-to-noise ratio in the experimental system,a fluorescence endoscope image acquisition system was designed and built.In the experimental system,the white images and the fluorescence images without the light source modulation and digital lock-in algorithm processing and with the light source modulation and digital lock-in algorithm processing are collected,respectively.And then the image signal-to-noise ratio is compared and analyzed.The experiment results show that the signal-to-noise ratios of white images and fluorescence images with digital lock-in algorithm processing are 39.54 dB and 35.70 dB,respectively.And the signal-to-noise ratios of white images and fluorescence images without digital lock-in algorithm processing are 36.56 dB and 33.47 dB,respectively.Compared with the image signal-to-noise ratio without digital lock-in image algorithm processing,the signal-to-noise ratio of the white images and the fluorescence images with digital lockin algorithm processing are increased by 8.2% and 6.7%,respectively.The experimental results in this paper show that the light source modulation technology and digital lock-in algorithm to demodulate white images and fluorescence images can enhance image signal-to-noise ratio,improve the image quality,more accurately analyze the fluorescence intensity distribution and accurately judge the position,size and shape of the tumor.So the fluorescence endoscope based on digital lock-in algorithm can improve the success rate of early stage cancer detection and assist the surgeon accurately identify and precisely remove the tumor.