摘要
目的:分析动态光谱法用于人体血液胆固醇含量无创检测的应用意义。方法:选取2013年10月-2014年1月本院96名健康体检者作为研究对象,对其进行在体测量动态光谱和抽血分析胆固醇含量,之后选用BP神经网络对获取的动态光谱数据和胆固醇含量实测值进行建模分析。结果:从96个体检样本中抽取81个作为校正集,另外15个作为预测集,得到校正集和预测集的相关系数分别为98.8%和95.65%,预测集的绝对误差最小为0.39%,最大为25.33%,平均相对误差为13.49%。结论:动态光谱法对胆固醇含量进行测量是一种精确性很高的无创检测分析方法,值得临床上推广。
Objective：To analyze the application significance of dynamic spectrum method for noninvasive detection of human blood cholesterol content.Method：96 healthy volunteers from October 2013 to January 2014 in our hospital were selected as research subjects,all volunteers received dynamic measurement of spectral and the cholesterol content analysis in the blood.After used BP neural network,the dynamic spectrum data and cholesterol content measured value were analyzed through the establishment of model.Result：81 patient samples extracted as the calibration set from 96 patient samples,the others as the prediction set,the correlation coefficient of the calibration set and prediction set were 98.8% and 95.65%,the absolute error of prediction set minimum was 0.39%,maximum was 25.33%,the average relative error was 13.49%.Conclusion：Dynamic spectrum method for cholesterol measurement is a high accuracy noninvasive detection analysis method,is worthy of clinical promotion.
出处
《中国医学创新》
CAS
2014年第15期136-138,共3页
Medical Innovation of China
关键词
动态光谱法
胆固醇
无创检测
Dynamic spectrum method
Cholesterol
Noninvasive detection
