论文部分内容阅读
目的:结合医用电子鼻技术,探讨糖尿病患者及其口腔呼气的气味图谱特征。方法:选择180例糖尿病患者和100例健康者,用医用电子鼻采集280例口腔呼气的气味图谱,采用基于数据特征划分的方法,用支持向量机和随机森林集成模型对糖尿病患者进行分类预测。结果:1线性核函数的支持向量机(SVM1)分类结果不是很理想,低于多项式核(SVM2)、径向基函数核(SVM3)和随机森林(RF)3种分类器,说明分类超平面显然是非线性的;2集成分类器对糖尿病患者和健康者的气味图谱特征的识别准确率可达88.04%。结论:基于特征划分的分类器集成方法预测性能明显好于单一分类器,为使用医用电子鼻进行糖尿病诊断分析提供了一种有效手段。
OBJECTIVE: To investigate the characteristics of the odor profiles of diabetics and their oral breaths in combination with medical electronic nose technology. Methods: 180 patients with diabetes and 100 healthy subjects were enrolled. 280 patients with oral breath breath pattern were collected by medical electronic nose. Based on the data feature classification method, the classification and prediction of diabetes patients by using support vector machine and random forest integration model . Results: The SVM1 classification results of a linear kernel function are not very good, which is lower than the SVM2, SVM3 and Random Forest (RF) classifiers, indicating that the classification hyperplane Obviously, it is non-linear. The recognition accuracy of 2 integrated classifiers can be up to 88.04% for the characteristics of odor profiles of diabetics and healthy subjects. Conclusion: The classification performance of classifier based on feature classification is obviously better than that of single classifier, which provides an effective method for the diagnosis of diabetes using medical electronic nose.