基于无创血糖检测技术的数据处理算法

Algorithm of Data Processing Based on Non-invasive Glucose Measuring

无创血糖检测仪的研究,对于糖尿病的治疗具有重要意义。该研究的原理是基于血糖是人体代谢的主要能源物质的能量代谢守恒法,通过测量人体指部的代谢率、血氧饱和度、脉率、血流量来估算血糖水平。由辐射散热、对流散热、蒸发散热量来确定代谢率,由热流温度场分布原理来测量血流量。最后通过多元回归分析的方法,建立血糖浓度的数学模型。经前后2组共30例临床初步实验,表明测试结果与AMS-AUTOLAB18全自动生化分析仪所测结果的相关系数达到了0.862 1,具有很好的相关性和一致性。同时,经误差统计分析得到检测结果的平均相对误差仅为0.039,平均相对误差界ε<0.5,表明整数部分均为有效数字位,且无显著性差异。多参数相关性分析结果表明,代谢率对血糖值的贡献是最大的,其相关系数达到了0.92,其他参数对血糖值的相关系数依次减弱。实验证明,所建立的基于能量代谢守恒法来无创伤地间接检测血糖值的方法是可行的。

Noninvasive blood sugar measuring apparatus is very important for diabetes. Based on the theory that blood sugar is the main energy resource, the method of conservation of energy metabolism （ COEM） estimates the blood sugar level with the metabolism rate, oxygen saturation, pulse frequency and blood flow of the part of finger. The metabolism rate was enumerated by calculating the heat diffuses of radiation, convection and evaporation. Besides, the method of temperature field distributing of heat liquid was used to measure the blood flow. Finally, the arithmetic model of the blood sugar concentration was constituted by regression analysis. The results of 30 preliminary clinical experiments showed that the correlative coefficient was 0. 8621 between the measured dada form noninvasive apparatus based on COEM and that from a automatic biochemical analyzer （AUTOLAB18） in the clinic. The average relative difference was 0. 039 and the boundary of average relative difference was less than 0.5. Besides, the metabolic rate was more important than other parameters using the correlative analysis. The correlative coefficient was 0. 92 between the metabolic rate and the blood sugar level. It was suggested that the proposed method was feasible to measure blood sugar.