一种测量动脉血氧饱和度的新方法

发布时间：2018-01-12 17:40

本文关键词：一种测量动脉血氧饱和度的新方法　出处：《光谱学与光谱分析》2016年01期 　论文类型：期刊论文

【摘要】：在传统的血氧饱和度测量方法中,首先计算对数光电容积脉搏波交、真流成分的比值(R值),然后通过实验建立血氧饱和度值与R值的线性回归模型,旨在消除光程的影响,将个体差异的影响降至最小化,是一种基于Lambert-Beer定律的降维处理。当考虑散射时,这种以交、直流成分的比值作为自变量的降维处理方式损失了信息量,引入了系统误差,限制了测量的精度。为了减小这种影响,从透射光电脉搏波所含信息量的角度,提出了使用双波长对数光电容积脉搏波的峰值和谷值作为自变量与血氧饱和度值建立线性回归模型的新方法。从23名健康志愿者中获得133例光电脉搏波及血氧饱和度参考值样本,从中随机抽取90例作为建模集,另外43例作为预测集,对比了新方法和R值法的建模和预测结果。随机抽取建模集并预测10次,新方法和R值法的平均预测相关系数分别为0.890 6和0.846 8,平均预测均方根误差分别为0.889 6%和1.037 3%,结果表明新方法的预测效果优于传统R值法,使用四个特征量建立血氧饱和度预测模型,能够提高测量的稳定性和精度,对基于有限波长光谱数据建模测量人体血液生理参数有指导意义。
[Abstract]:In the traditional blood oxygen saturation measurement method, the logarithmic photovolumic pulse wave intersection and the ratio of true current components are calculated firstly, and then a linear regression model between the oxygen saturation value and the R value is established through experiments. Aiming at eliminating the influence of light path and minimizing the influence of individual difference, it is a dimensionality reduction process based on Lambert-Beer 's law. The ratio of DC component as the dimension reduction method of independent variable has lost information, introduced systematic error, and limited the accuracy of measurement. In order to reduce this effect, the angle of information contained in transmitted optoelectronic pulse wave is introduced. A new method for establishing a linear regression model using the peak and valley values of logarithmic photovolumic pulse wave as independent variables and the value of blood oxygen saturation was proposed. 133 healthy volunteers were selected as photovoltaic pulse wave waves. Blood oxygen saturation reference value sample. 90 cases were selected as modeling set and 43 cases as prediction set. The modeling and prediction results of the new method and R value method were compared. The average predictive correlation coefficients of the new method and the R value method were 0.8906 and 0.8468, respectively. The mean root mean square error (RMS) of the new method was 0.889 6% and 1.037 3%, respectively. The results show that the prediction effect of the new method is better than that of the traditional R value method. Using four characteristic variables to establish the oxygen saturation prediction model can improve the stability and accuracy of the measurement. It is of guiding significance for modeling and measuring human blood physiological parameters based on finite wavelength spectral data.
【作者单位】：天津大学精密测试技术及仪器国家重点实验室;天津大学天津市生物医学检测技术与仪器重点实验室;华东师范大学极化材料与器件教育部重点实验室;天津市人民医院检验学部;天津市人民医院查体中心;
【基金】：国家自然科学基金项目(30973964) 天津市应用基础及前沿技术研究计划项目(11JCZDJC17100)资助
【分类号】：R446.11
【正文快照】： 1.天津大学精密测试技术及仪器国家重点实验室,天津3000722.天津大学天津市生物医学检测技术与仪器重点实验室,天津3000723.华东师范大学极化材料与器件教育部重点实验室,上海2002414.天津市人民医院检验学部,天津3001205.天津市人民医院查体中心,天津300120引言血氧饱和度是

【相似文献】