基于人体呼出气体标志物检测的口腔疾病筛查与诊断研究

Study on the Screening and Diagnosis of Oral Diseases Based on Volatile Sulfur Compounds in Human Exhaled Breath

建立基于气相色谱与质谱(GC-MS)联用技术分析口腔内挥发性硫化物(VSCs)浓度的标准化方法和流程。利用气袋采集65个受试者的呼气样本,并采用GC-MS和临床使用的口气检测仪分析样本内VSCs浓度,对两种方法检测的浓度值进行相关性分析。进一步采用ROC法建立口臭等级阈值模型,设计基于线性判别分析、逻辑回归、支持向量机等多种方法的牙周疾病诊断模型。结果表明,GC-MS法与呼气检测仪检测的VSCs浓度值具有相关性(P<0.01)。基于ROC的口臭等级阈值模型对口臭等级区分的灵敏度可达到0.897,基于逻辑回归的牙周疾病诊断模型的灵敏度和特异度分别达到0.733和0.771。所提出的基于口腔内VSCs筛查和诊断口腔疾病的方法是可行的,进一步结合其他呼出气体标志物同时检测,有望为疾病诊断提供更简便、无创和快速的手段。

This study established a standardized method and procedure to detect volatile sulfide compounds(VSCs)in human exhaled breath by gas chromatography and mass spectrometry(GC-MS)for the first time.By this way,65 breath samples were collected using airbags,and the concentrations of VSCs were analyzed by GCMS and instruments currently used in clinical applications,the correlation of which were analyzed.Further,receiver operating characteristic curve(ROC)method was adopted to establish the threshold model of halitosis grade,and the periodontal disease diagnosis models based on linear discriminant analysis,logistic regression,and support vector machine were designed.The results of GC-MS method were correlated with the breath detection instruments(P<0.01).The sensitivity of halitosis grade threshold diagnostic model based on ROC reached 0.897 at present.The sensitivity and specificity of the periodontal disease diagnosis model were 0.733 and 0.771 respectively by logistic regression method.This work validated the feasibility to screen and diagnose oral diseases based on VSCs with good performance.It could be predicted that the combination with other markers of exhaled gas would provide a more simple,non-invasive and rapid diagnostic way for disease diagnosis.