Research On The Technologies Of Oral Cancer Cell Image Recognition

With the development of modern society in the world, the incidence of oral cancer each year increased. As a worldwide census results showed that in all the tumor lesions in the mouth, throat cancer in males accounted for fifth place in the seventh place of women in the total population of the sixth. Oral cancer is a deadly disease (relative risk of lip cancer less), the 5-year survival rate is only 30-40% (lip cancer is about 80 per cent). Despite the continuing efforts of professionals, modern technology has greatly improved the treatment of patients with malignant tumors of the oral quality of life, but on the survival rate of such patients, the past several decades has not been significantly improved. This is mainly because, despite the original malignancy itself to be an effective treatment, however, patients may suffer cancer lesions of the invasion and metastasis of cancer deaths. Therefore, oral disease especially in the early oral cancer prevention, early detection and treatment, the cure rate for improving disease, radical tumor is of great significance.Pathologic diagnosis of tumor in the course of the general morphology of observation and qualitative description. Research on cell morphology and the most important way is thin slice under a microscope, the microscope images obtained by the traditional method mainly in the large number of observations on the basis of the qualitative description. This approach, on the one hand inspection workload and low efficiency test, physicians work fatigue or careless mistakes caused identification, impact of disease diagnosis; on the other hand, pathological image analysis to identify vulnerable physician experience and visual resolution restrictions , prone to subjective factors, the lack of scientific basis for the quantitative objective.With quantitative image analysis and recognition technology in the development of the extensive application of Pathology, Pathology of a new branch of computer image recognition, image recognition technology will be introduced to doctors in the diagnosis process, the application of computer image analysis technology to identify research The pathology of the disease and structure characteristics of their diagnosis, classification, classification and prognosis judgement in the application, and so on. Can enhance recognition accuracy, but also to reduce the labor intensity and workload as a result of the elimination of the psychological adaptation, as well as disadvantages, such as fatigue caused by misdiagnosis and missed diagnosis can be a very large extent, help doctors correct diagnosis .Oral Buccal cell changes in the pattern of the early days, always volume changes and objectively accurate to the specific data used to describe such a quantitative change, accurate description of the morphology of the cells in different physiological and pathological conditions to conduct an objective comparison, it is necessary to carry out quantitative analysis of cell morphology. And through quantitative analysis of this study, certain lesions can also find the law, to improve the accuracy of diagnoses provide the basis for the rate. And image analysis technology for such microscopic image of quantitative analysis provides a powerful tool. The topic is to use image processing and pattern recognition technology to oral tissue biopsy microscopic image implementation of quantitative analysis, research for oral pathology images of the cells in the classification, measurement, recognition algorithm, for the correct diagnosis of medical pathology provide objective data.In this paper, mathematical morphology, image processing and pattern recognition theory, the image edge detection, in view of cancerous cells and normal cells little difference in the characteristics of the cancer cells to recognition algorithm. With the traditional classic edge detection method, the method used in this paper better able to eliminate noise interference, and more accurate extraction of the characteristics of continuous edge. Choose accurate, and appropriate description of the characteristic parameters of this image is in focus.The main research work mainly in the following aspects: First of all, a picture of the pathological pretreatment. Adopting a gray histogram equalization law, passed quickly in the value of image segmentation identification before processing. Then, will be widely applied to the analysis of the binary image to promote mathematical morphology gray image computing, using morphological description of the signal shape, scale characteristics of the advantages of using morphological image filtering algorithms, accurate extraction of the target. To the adoption of neural network training, and the classification of design optimization programme.Finally, the basic principles of visual detection of oral cancer Image Recognition System for the diagnosis were introduced. On the image acquisition system design, image preprocessing cells, the cell extraction, as well as the nucleus of cells filled contour tracking, and other related aspects of in-depth studies, and achieved good results. The experiments show that the system can be achieved for the diagnosis of oral cancer cells function, we can alleviate the pressure of medical personnel, automated testing, the census work for the promotion of cancer.