Development Of Endoscopic Cerenkov Luminescence Imaging System And Its Dynamic Quantitative Software

Gastric cancer seriously threatens people’s life and health.Its special lesion location leads to difficult diagnosis,complicated surgery and low postoperative survival rate.The current research programs for gastric cancer mainly focus on early detection and personalized treatment.Endoscopy-based diagnosis is based on the morphological changes of the tumor,but early gastric cancer is difficult to distinguish,making accurate early diagnosis difficult.During the development of gastric cancer,the change of molecular function of tumor cells is earlier than morphological changes.Therefore,the development of imaging technology that can identify changes in molecular levels of tumor cells is very important for the early diagnosis of gastric cancer.Combined with the emerging Cerenkov luminescence imaging(CLI)technology,endoscopic Cerenkov luminescence imaging(e CLI)can achieve gastric cancer-specific functional imaging at the cellular and molecular level.However,the signal acquisition efficiency of the existing Cerenkov luminescence endoscopes(CLEs)is too low,which makes it unable to meet the needs of clinical use.On the other hand,the occurrence of cancer usually results in abnormal expression of cell surface receptors,so accurately quantifying the availability of cell surface receptors is of great significance for the accurate diagnosis,personalized treatment,development of antitumor drugs,and evaluation of drug efficacy.However,molecular probes that specifically bind to receptors have particularly complex pharmacokinetic properties,making it difficult quantify receptors on cell surfaces in vivo.Therefore,there are two major problems in early accurate diagnosis and treatment of gastric cancer based on e CLI technology,including the low efficiency of signal acquisition of the existing CLEs,and the difficulty in achieving in vivo quantification of tumor receptors.This paper focuses on these two issues of e CLI technology,and mainly conducts the following two aspects of research:First,in order to solve the problem of low luminescence signal collection efficiency of CLE,this paper optimizes the system structure from the hardware level,so as to reduce the loss of the Cerenkov luminescence signal during the use of endoscope.Firstly,the existing CLE structure was analyzed,and factors affecting the signal collection efficiency were summarized,including the monofilament diameter of the fiber imaging bundle of the endoscope,the field-of-view angle of the fiber imaging bundle probe,and the materials of fiber imaging bundle,etc.Secondly,by comparing the imaging effects of different endoscope adapters,the monofilament diameters of the fiber imaging bundle,the field of view of the fiber imaging bundle probe,and the material of the fiber imaging bundle,we determined the system structure with the optimal parameters.Finally,CLE was equipped with these optimal parameters,and was tested for the spatial resolution and detection sensitivity,including the white light and luminescence spatial resolution,in vitro and in vivo sensitivity.In addition,this article also explored the enhancement effects of the commonly used scintillators on Cerenkov luminescence signal as a reference for clinical use of endoscopes.Second,in order to accurately quantify the receptors on the surface of gastric cancer cells in vivo,this paper explores a method for quantifying receptors based on dynamic e CLI technology and develops a tumor receptor quantification software platform.The software platform in this paper mainly includes the functions of processing dynamic optical images,circling the region of interest(ROI),extracting the time activity curve(TAC),and solving the dynamic models.Firstly,batch processing was performed on the acquired sequence dynamic luminescence images,including distortion correction of endoscopic images,highenergy ray noise denoising,and background removal.Secondly,the ROI of the sequenced dynamic fluorescence image was circled after batch processing,and the average intensity of luminescence signal in the ROI was calculated to extract the TAC,correct radionuclide decay attenuation,and perform data management and display.Finally,a compartment model was selected for solving the extracted TAC.The selectable compartment models mainly include logan graphical analysis with the reference tissue model,simplifies the reference tissue model and Gurfikel exponential model.The overall framework and most of the functions of the software platform are implemented using Qt and C++.The model calculation part uses the mixed programming techniques of C/C++ and MATLAB.