Study On The Differentiation Of ? And Non ? Phase Of Ovarian Endometrioma

?Objective? To explore the effective methods to distinguish stage ? of ovarian endometrioma patients and non ? stage(stage I-III)of patients by modeling methods.At the same time,make a preliminary assessment of NLR and combined marker(NLR * CA125)in the diagnosis of stage ? ovarian endometrioma patients,Therefore they can serve the clinical diagnosis and treatment better.?method? Our study adopted the method of retrospective study,ovarian endometrioma patients receiving surgery in our hospital from January 2015 to december 2016 were included and all patients were confirmed by pathological examination.We adopted the r ASRM staging method to distribute them into stage I-?,then we selected 40 cases of patients with non ? stage,as well as 35 cases of patients with stage ? as the research object.17 indicators including age,CA125,HE4,cyst area,white blood cell count,neutrophil percentage,rate of lymphocytes,monocyte ratio,alpha hydroxybutyrate dehydrogenase,cholesterol,triglyceride(TC),low density lipoprotein(LDL),high density lipoprotein(HDL),lactate dehydrogenase(LDH),alkaline phosphatase(ALP)were selected as descriptors.We selected the best combination of these indicators to create the final model by using the method of support vector machine model.Supposing that the patients with stage ? were positive for diagnosis,then use ROC curve to evaluate the sensitivity and specificity of the indicator selected,NLR,CA125,and combined marker in the diagnosis of stage ? patients of ovarian endometrioma.Chi-squaretest was used to compare the the positive rate of NLR,CA125 and combined maker between stage ? patients and non ? stage patients,and evaluate their correlation with staging.?results? The results of ROC curve showed that when used alone,the ability of 8 indicators including age,CA125,HE4,cyst area,LDH,LDL,HDL and ALP that we selected to diagnose stage ? patients of ovarian endometrioma were poor,(Area under the curve of ROC < 0.7).Wheres,when they work together by the support vector machine,the sensitivity and specificity increased significantly(82.8% and 77.5%).At the same time,the result of chi-square test showed that NLR doesn't correlate with stage of ovarian endometrioma(P > 0.05),but CA125 and combined maker are correlated with stage of ovarian endometrioma(P < 0.01 and P<0.05).The results of ROC curve indicates that the diagnostic ability of these three indicators above in stage ? patients of ovarian endometrioma is not so high(AUC < 0.7).?Conclusion?(1)The ability of these indicators above to diagnose stage ? ovarian endometrioma patients were poor.Sensitivity and specificity improve significantly when combine several indicators;(2)NLR,CA125 and combined marker have poor ability to diagnose the stage ? patients of ovarian endometrioma.