Analysis And Application Of Tongue Features In High Risk Population Of Esophageal Cancer

BackgroundCixian was a high incidence area of esophageal cancer in China, the incidence of esophageal cancer was 91.5/100,000, significantly higher than the national average.2005 Cixian began early detection and treatment of esophageal cancer project, endoscopic census for the local high-risk population (40-69 years old), histopathological examination for suspicious lesions. However, the incidence of local esophageal cancer has not been effectively controlled. Chinese medicine belived that tongue connected through meridians to internal organs, organ conditions change, the properties and functions can be found by looking at the tongue, tongue diagnosis of traditional Chinese medicine in the role of esophageal cancer screening was worth in-depth study.Parti:esophageal cancer high-risk groups Characteristics of TongueObjectiveTo analysis the differences of tongue between different pathological diagnosis of esophageal cancer high-risk groups, including the tongue color space values and tongue feature.MethodsThree thousand and fifty-three cases were included in Cixian with a high incidence of esophageal cancer, it were divided into normal group, esophageal low grade intraepithelial neoplasia group, esophageal high-grade intraepithelial neoplasia group according to the pathological diagnosis and gastroscopy. Besides, basic information was collected, tongue color space color values (including HSV, Lab, RGB color spaces) and tongue feature (including tongue color, tongue color, tongue body fa tor thin, a total of 10)Results1. Sex and age between normal, esophageal neoplasia low-level group, the high-level inter-esophageal neoplasia group showed statistically significant differences (p< 0.01).2. In the HSV color space, the comparison of S-value between the esophageal high grade intraepithelial neoplasia group and the normal group, showed statistically significant differences (P< 0.05).3. In the Lab color space, the comparison of b value between the normal group and the low grade+high grade intraepithelial neoplasia groups, showed statistically significant differences (P<0.05).4. The overall distribution of tongue ecchymosis between normal esophageal neoplasia low-level group, the high-level inter-esophageal neoplasia group showed statistically significant differences (P<0.01), the proportion of tongue ecchymosis of esophageal neoplasia low-level group and the high-level inter-esophageal neoplasia group was higher than the normal group.5.The overall distribution of tongue color between inflammation group and no n-inflammatory group showed statistically significant differences (P<0.01).6. The overall distribution of tongue ecchymosis between inflammation group and esophageal cancer and precancerous lesions group showed statistically significant differences (P<0.01).ConclusionThe tongue feature between normal, esophageal neoplasia low-level group, the high-level inter-esophageal neoplasia group showed statistically significant differences. tongue ecchymosis showed statistically significant differences.Part II:Establishment and verification based Tongue esophageal cancer and precancerous lesions logistic regression modelObjectiveTo analyze the risk factors for esophageal cancer and precancerous lesions by building precancerous lesions and carcinoma logistic regression model with tongue feature, gender and age, and to explore the cause and pathogenesis of esophageal cancer, at the same time, to carry on application of esophageal cancer screening model.MethodsThree thousand eight hundred and eighty-three cases were included, according to the time of enrollment, divided into training group (3053 cases) and test group (a total of 830 cases). The tongue feature, sex and age of training group were assigned to establish esophageal cancer and precancerous lesions logistic regression model, and the accuracy of the model was verified by detecting the group.Results1. Logistic regression analysis showed that age, male, red tongue, tongue petechiae, yellow and white moss phase and were risk factors for esophageal cancer.2. Internal diagnostic specificity of logistic regression model was 72.80%, a sensitivity of 61.60%, the total coincidence rate was 72.3%. External diagnostic specificity of 80.35%, a sensitivity of 63.41%, the total coincidence rate was 79.51%.ConclusionThe logistic regression model of esophageal cancer based on Tongue, gender, age found that features tongue red tongue, tongue petechiae, yellow and white moss phase wre risk fectors for esophageal cancer and precancerous lesions, it was used to test esophageal cancer high-risk groups, and had a high coincidence rate between pathological diagnosis.