The Detection Of Oral Cancer And Tongue Squamous Cell Carcinoma Animal Model By Fiber Raman Spectroscopy

Background:Oral cancer is a common clinical disease.In recent years,its incidence is on the rise,which seriously threatens the safety of patients.The current treatment of oral cancer is a comprehensive treatment based on surgery.Early diagnosis of the disease directly affects the prognosis of patients.Raman spectroscopy is a technique that uses optical-substance interactions to produce inelastic scattered light for optical detection.It can detect biochemical and biomolecular structures and tissue conformations,providing"chemical fingerprint information" for cells,tissues or biological fluids.With the development of related technologies and optical instruments,Raman spectroscopy has become a research hotspot in the medical field in recent years,and it is expected to differ from optical diagnosis of cancer tissues and normal tissues.Based on fiber-optic Raman spectroscopy,combined with convolutional neural network and support vector machine analysis algorithm,this study detects surgically resected oral cancer tissue,adjacent normal tissues,establishes CAL-27 tongue squamous cell carcinoma animal model,and then grew tongue squamous cell carcinoma.Obtain the Raman spectrum of tumor tissue and contralateral normal tissue with the Raman System.Finally analysis the spectrum.It mainly includes the following three aspects:1.Detection of oral cancer tissue based on fiber Raman spectroscopy Objective:Using fiber Raman spectroscopy to detect Raman spectra of oral and cancerous tissues.Classifying cancerous tissues and cancer based on fiber Raman spectroscopy.Methods:Surgical resection of oral cancer tissue and normal adjacent tissue specimens confirmed by frozen pathological results were performed.Fiber Raman spectroscopy was performed,smoothing algorithm and de-baseline algorithm were used to process spectral images,and convolutional neural network was used to support the data.Vector machine establishes a deep learning model for classification of tongue squamous cell carcinoma.Results:The cancer tissue and the normal tissue were in the same coordinates,and the Raman spectrum was more convenient and intuitive.The spectrum of the cancer tissue and the adjacent normal tissues showed the same trend,and there were differences in individual peaks and peaks.266 tongue squamous cell carcinomas Cross-validation with the test set and training set of the adjacent normal tissue spectrum,one spectral classification error.Conclusion:Fiber Raman spectroscopy can be used to distinguish between oral cancer tissues and adjacent normal tissues.Combined with convolutional neural networks and support vector machines,the specificity and sensitivity are high.2.CAL-27 tongue squamous cell carcinoma tumor model establishment Objective:To establish a CAL-27 tongue squamous cell carcinoma tumor model for fiber Raman spectroscopy.Methods:The CAL-27 tongue squamous cell carcinoma cell line was cultured,and the logarithmic growth phase cells were selected.Six 5-6 week old female nude mice were selected twice,and inoculated into the nude mice at a concentration of 1 x 106 cells/mL.Then,the cancer tissue and normal tissues were taken for pathological examination.Results:After 1 week,it was observed that there were about miliary size tumors at the inoculation site,and the tumors could be used for 3 to 4 weeks.The first tumor formation was 2/6(33.3%),and the second was 5/6(83.3)·Conclusion:The method for implanting tumors of tongue squamous cell carcinoma can establish a tumor model relatively stably.3.Detection of tongue squamous cell carcinoma model in nude mice based on fiber Raman spectroscopy Objective:To detect the in vivo based on fiber Raman spectroscopy by detecting the cancer tissue and contralateral normal tissue spectrum of nude mice in CAL-27 tongue squamous cell carcinoma model.Cancer tissue and normal tissue.Methods:The animal model of tongue squamous cell carcinoma was sacrificed,and the Raman spectroscopy of the cancer tissue and normal tissue was performed in vivo.The smoothing algorithm and baseline algoritum and normalization algorithm were used to process the spectral image.Results:The spectrum of cancer tissues and normal tissues was obtained.The spectrum of cancer tissues and normal tissues showed the same trend.The normalized spectrum of cancer tissues and normal tissues showed the same trend of the spectral lines between cancer tissues and normal tissues.Conclusion:Fiber Raman spectroscopy can distinguish tumor and normal tissues from tongue squamous cell carcinoma.