A Model Prediction Study On Vaginal Microecological Index And High Risk HPV Infection

Background and Objective:Current studies indicate that persistent infection of high-risk human papilloma virus(HPV)is the main cause of cervical intraepithelial neoplasia and cervical cancer.Recent studies have shown that vaginal microecological changes play a role in the development of high-risk HPV infections,cervical intraepithelial neoplasia and cervical cancer.However,the correlation between vaginal microecological changes and high-risk HPV infection has not been unified,and the specific indicators in vaginal microecology.This paper mainly uses machine learning algorithm to try to construct the prediction model of vaginal microecological index and cervical high-risk HPV infection risk,in order to explore the risk factors of high-risk HPV infection,so as to provide theoretical basis for preventing and interfering with high-risk HPV infection?Methods:This study analyzed 1135 patients in our hospital from august 2018 to june 2019,who underwent vaginal microecological examination and HPV detection the information of vaginal microecological examination and infection of high-risk HPV were collected.Using spss24.0 software to analyze the data,to compare the infection rate of high-risk HPV of all ages with the infection rate of various vaginitis using chi-square test,p<0.05 thought the difference was statistically significant;using chi-square test and fisher exact probability method to analyze the indicators of vaginal microecology and high-risk HPV infection by single factor analysis,p<0.05 thought the difference was statistically significant.the random forest model in machine learning was also used to learn the correlation between vaginal microecological indicators and high-risk HPV infections.the validation set was then substituted into the model to calculate the error rate to a minimum and to predict the contribution value of a single variable at the same time.the effectiveness of the model was evaluated by the subcurve area(AUC)of the subjects' work characteristics(ROC).General analysis of patients enrolled1.Distribution of high-risk HPV infections by age groupIn the 31-40 age group,the infection rate of high-risk HPV was 27.02%,followed by 24.14% in 41-50 years old,and the distribution of high-risk HPV infection was statistically significant(p<0.05).2.Analysis of Vaginal Microecology in PatientsVaginal microecological imbalance accounted for 84.76% of the total in 1135 patients,298 cases of definite infection of vaginal microecological imbalance accounted for 26.26% of the total,664 cases of unidentified infection of vaginal microecological imbalance accounted for 58.50% of the total.Analysis of vaginal microecological index and single factor of high-risk HPV infection.Single infection accounted for 81.21%,double infection 18.12%,triple infection 0.67%,the vulvovaginal candidiasis(VVC)infection rate was 37.58%,bacterial vaginosis(BV)was 30.87%,aerobic vaginitis(AV)was 30.87%,trichomonas vaginitis(TV)was 1.68%.The most common in VVC and BV,higher than other vaginitis rates.Analysis of vaginal microecological index and single factor of high-risk HPV infection1.Analysis of vaginal microecological function index and single factor of high-risk HPV infectionThe detection rates of PH>4.6,hydrogen peroxide,leukocyte esterase and sialidase in the HPV positive group were 60.31%,34.13%,67.06% and 20.63%.The detection rates in the negative group were 51.07%,26.95%,57.53% and 14.84%,and the difference was statistically significant(p<0.05).2.Analysis of microecological morphologic index of vagina and single factor of high-risk HPV infectionThe detection rate of BV?BV?VVC? heteromycetes and abnormal density of flora in HPV positive group was 15.87%,8.33%,17.06%,63.10% and 12.70%.The detection rate was 10.31%,4.87%,12.00%,53.00% and 6.8% in HPV negative group(p<0.05).A Risk Prediction Model Based on Microecological Changes of Vagina and high-risk HPV of InfectionBV is the most important of the related factors of infection with high-risk HPV,which is positively related to infection,which is the high-risk factor leading to high-risk HPV infection.the second was leukocyte esterase,AV,VVC,PH increase,decreased lactobacillus,hydrogen peroxide deficiency,sialidase and high-risk HPV infection.However,the model did not point out TV,there was a significant correlationResults: between bacterial density and diversity and high-risk HPV infection.By deriving the prediction model of risk of high-risk HPV infection and the working characteristics of subjects,the area under the curve was 0.962,and its 95% confidence interval was(0.919-0.986).Conclusions:1.BV?AV?VVC? lactobacilli decreased,PH >4.6,hydrogen peroxide decreased,leukocyte esterase,sialic anhydrase were risk factors for high-risk HPV infection,among which BV was the most important factor for high-risk HPV infection.2.Vaginal microecological imbalance is related to high-risk HPV infection,combined with vaginal microecological information can establish a risk model to predict high-risk HPV infection,and have a good predictive effect.