Development of a New Type of Electronic Nose System Equipped with Metal Oxide Sensors

本研究尝试以预浓缩器与六金属氧化物半导体感测器阵列相结合的新型电子鼻来评估恶臭气体。受试者人群包括55名主诉口腔恶臭的患者。采用气相色谱法测定挥发性硫化物（VSC）浓度，感官试验（奥尔特）和预测奥尔特评分（EN-OL 1）评价口腔恶臭。使用新型电子鼻系统，通过多元线性回归方法估计EN-OLT评分作为恶臭强度。EN-OLT与奥尔特评分的相关系数（r=0.71）高于VSC与奥尔特评分的相关系数（r=0.71）。以性别、年龄、吸烟、饮酒、PPD、TCS和PCR为自变量进行线性和Logistic回归分析。在线性回归分析中，这些独立变量的解释能力显示EN-OLT评分为37%，GC-VSC为16%，奥尔特评分为32%。在以受试者位于EN-OLT评分分布的第25百分位数的可能性作为因变量的logistic回归分析中，PPD、TCS和PCR显示出显著相关性，其比值比分别为17.1、12.4和28.0。总之，口腔恶臭强度测量与电子鼻可能是有用的，用于评估临床口腔恶臭，并可能与牙周状况，舌苔和口腔卫生状况。

In this study, we attempt to evaluate malodor with a new type of electronic nose which combined with the preconcentrator and six metal oxide semiconductor sensor array. The subject population consisted of 55 patients who complained of oral malodor. Oral malodor was assessed a concentration of volatile sulfur compaunds (VSC) measured by gas chromatography, organoleptic test (OLT) and predicted OLT score (EN-OL1). EN-OLT score was estimated as malodor strength by the multiple linear regression method using a new type of electronic nose system. The correlation coefficient between EN-OLT and OLT score (r=0.71) was higher than that between VSC and OLT score. Gender, age, smoking, alcohol consumption, PPD, TCS and PCR as independent variables were used in linear and logistic regression analyses. In linear regression analyses, the explanatory power of these independent variables showed 37% for EN-OLT score, 16% for GC-VSC and 32% for OLT score. In logistic regression analyses on possibility of subjects being in upper 25th percentile of EN-OLT score distribution as a dependent variable, PPD, TCS and PCR showed significant associations and their odds ratios were 17.1 and 12.4 and 28.0, respectively. In conclusion, oral malodor strength measured with the electronic nose may be useful for evaluating clinically oral malodor and may be associated with periodontal status, tongue coating and oral hygiene status.