A logistic regression model based on tongue image information for prediction precancerous lesions and early stage esophageal cancer in China

Background

China is an esophageal cancer high incidence country, with more than 50% esophageal cancer case of the world. Screening and diagnosis of precancerous lesions and early stage cancer are main measures of decreasing incidence rate and mortality rate.

Methods

High-risk population (40-69 years old) in Cixian a high risk area of esophageal cancer was screened, and 3053 cases were included. They were devided into 3 groups: normal group, esophageal neoplasia low-level group and esophageal neoplasia high-level group, according to pathology and electronic gastroscope diagnosis. Diagnostic testing lingual information collection system (DS01-B) was used and tongue image were collected, including tongue color, coating color, fur character, tongue shape, local ecchymosis, et al. Difference of tongue image information was analyzed, related clinical variants were analyzed by multi-factor logistic regression.

Results

Incidence of local ecchymosis in tongue image was 1.58% (45/2840) in normal group, 2.98% (4/134) in esophageal neoplasia low-level group and 6.32% (5/79)in esophageal neoplasia high-level group. Significant difference was found in 3 groups (P 

Conclusions

Red tongue, tongue local ecchymosis, yellow and white coated tongue were risk factors for precancerous lesions and early stage esophageal cancer. Multi-factor (including tongue image information) logistic regression model has clinical value of prediction precancerous lesions and early stage esophageal cancer.

Clinical trial identification

Legal entity responsible for the study

China-Japan friendshop hospital

Funding

National “twelfth five”science and technology support plan

Disclosure

All authors have declared no conflicts of interest.