Application Of Fractional Exhaled Nitric Oxide And Data Mining In Chronic Cough

PART Ⅰ Clinical diagnosis and preliminary analysis of chronic coughObjective:Patients with chronic cough account for a large proportion of outpatients.Analyze the data of various tests with chronic cough in order to invest the causes of chronic cough and the quality of life of patients.Methods:A total of 137 patients with chronic cough from April 2017 to March 2018were enrolled in this study.Body basic situation,routine blood test,pulmonary function test,bronchial provocation test（BPT）and questionnaire investigation were given.Results:70 patients had positive BPT result.They were diagnose Cough Variant Asthma（CVA group）finally.67 patients had negative BPT result（non CVA group）.There was a significant difference between the two groups in Forced Expiratory Volume in First Seconds（FEV_l）（92.751±15.508&102.897±13.180,P﹤0.001）,Forced Expiratory Volume in First Second/Forced Vital Capacity（FEV₁/FVC）（80.162±8.170&86.557±5.439,P﹤0.001）,Peak Expiratory Flow（PEF）（86.614±15.477&96.234±13.845,P﹤0.001）and Maximum Midexpiratory Flow（MMEF）（66.261±22.011&93.51±23.440,P﹤0.001）.There was a significant difference between the two groups in eosinophils count（EOS）（0.340&0.110,P﹤0.001）,eosinophils ratio（EOS%）（2.950&1.400,P﹤0.001）and mean corpuscular hemoglobin concentration（MCHC）（325.329±11.372&331.985±9.619,P﹤0.001）.Cronbach’sαof three fields in Leicester Cough Questionnaire（LCQ）are 0.623,0.911and 0.692.Gender differences might be exist.Conclusions:Female patients with chronic cough are more than male patients.Airway functions of all patients remain almost normal.There was a decline of airway functions in CVA group,but on the other hand there was an increase of EOS in CVA group.The research found MCHC had a significant decline in CVA group for the first time.Questionnaire investigations might reflect the quality of life in patiets.PART Ⅱ The diagnostic value of Fe NO in Cough variant asthmaObjective:Cough variant asthma（CVA）is the main cause of chronic cough.Analyze the data of Fe NO in order to invest its diagnostic value in CVA.Methods:A total of 137 patients with chronic cough from April 2017 to March 2018 were enrolled in this study.Fe NO test,routine blood test,pulmonary function test,BPT and questionnaire investigation were given.Results:There was a significant difference between CVA group and non CVA group in Fe NO（88 & 17,P﹤0.001）.All patients were divided into four groups by age.Fe NO decreased with the age increasing.Differences in four groups were significant（123.500±56.459 & 110.150±59.854 & 77.455±53.109 & 38.429±39.029,P=0.015）.EOS and EOS% in CVA patients were both positive correlation with Fe NO（r=0.509,P<0.001;r=0.465,P<0.001）.The area under the ROC curve（AUC）of Fe NO was 0.88（95%CI:0.81-0.95,P<0.05）,with an optimal cutoff value of 34.5ppb for maximum Youden index of 0.69.Cough symptom score in CVA patients was negative correlation with Fe NO（r=-0.64,P=0.046）.Conclusions:Elevated Fe NO level could be detected in CVA patients.Fe NO showed good correlation with EOS.Fe NO,as an inflammatory marker,might reflect airway inflammatory.It might be useful in the diagnosis and antidiastole of chronic cough.PART Ⅲ Analysis of cough detection data based on decision tree and support vector machineObjective:Using data mining,analyse cough detection data based on decision tree and support vector machine（SVM）.Methods:Modeling data was used by decision tree and support vector machine.Analyse the two methods in measuring the importance of detection index implementation and algorithm effectiveness.The platform of the following experiments is the Intel Core i7-5600 U CPU,12 GB memory and windows 7（64 bit）.The programming language is java.Results:Based on the establishment of two diagnostic data models,Fe NO and EOS% were important markers in the diagnosis of cough.At the same time,the experiment results found that the established models were effective for classifying the unpredictable test data.Conclusions:The research showed that the decision tree method and SVM had practical feasibility and effectiveness in diagnosis of cough.