| Liver disease is a commonly clinical disease,especially primary hepatic carcinoma that rapidly develops with high mortality and heavy medical burden,which brings heavy burden to patients and family members.The number of patients with liver diseases in the world is about 1.3 billion,including western 0.5 billion patients and 0.4 Chinese patients.It was reported by the International Agency for Research on Cancer(IARC).There are 782,000 new cases of liver cancers,of 50%which was contributed by China.In 2016,China’s annual statistical yearbook on health and family planning shows that the death rate of liver cancer in urban and rural residents in China is up to 23.95/10 million,ranking second in cancer mortality.Therefore,it is very crucial for patients to screen and intervene in the early stage of liver diseases.Alanine aminotransferase(ALT)and Alpha fetoprotein(AFP)are used to diagnose and screen liver,especially.The clinical use of alanine aminotransferase is often used in the diagnosis of liver disease,and the use of alpha fetoprotein to screen liver cancer.However,the studies primarily focus on biological characteristics of the indicators.this paper,thus,study the relationship between ALT and AFP values and geographical factors,and its spatial distribution characteristics to provide the basic tenet for clinical development.Based on the methods of literature searching and field investigation,the ALT and AFP values for healthy adults have been collected and sorted.The reference values of ALT were collected in 121977 cases from 93 units,67941 in men and 54036 in women.There were 9364 cases from 96 units of AFP values,5215 men and 4149 women.According to physical environment changes in China,4 categories of 22 geographical indicators are selected.Firstly,the correlation analysis was used to extract the geographical factors related to the two medical indexes,and the contribution rate of each factor was evaluated.Secondly,the methods of ridge regression and support vector machine were employed to established models of ALT and AFP.Furthermore,T-test and relative error analysis were used to evaluate the accuracy and validity of’ models.Thirdly,spatial auto-correlation,hot-spot analysis and trend surface were used to analyze the spatial characteristics of ALT and AFP values.Finally,we makes maps of ALT and AFP distribution with Kriging interpolation.Furthermore,we explore the mechanism of the effect of geographical factors on ALT and AFP values,and conclude the distribution laws.Through qualitative and quantitative studies,there are spatial disparities of the two values(P<0.05)and spatial auto-correlation characteristics(P<0.01).The correlation analysis showed that ALT values are related with 6 geographical factors.It is negatively correlated with the longitude,annual sunshine time,annual mean temperature,annual mean relative humidity and annual precipitation,and is positively related to the altitude.The AFP values are related to 7 geographical factors.The longitude,altitude,average annual temperature,annual precipitation,soil silt percentage and other factors were negatively correlated with latitude and temperature were positively correlated.the established model has a good fitting and be applied to predict the unknown sample points via T-test and relative error test.Finally,we draws maps of the two values.Through the above analysis,we can draw the following conclusions:(1)There are spatial disparities of ALT and AFP values that are related with geographical factors.The spatial distribution of values show western areas are higher eastern ares and northern areas are higher southern areas.(2)Linear model and nonlinear have own advantages and drawbacks.The higher precision of the nonlinear model which can effectively solve the problem of multiple collinearity and over fitting.(3)Drawing the distribution of two values for healthy adults,clearly shows the trend of changes in 3D space,which not only facilitate the represent of ALT and AFP values,but conducive to clinical application. |
PDF Full Text Request