基于拉曼光谱的血液遗留时间研究与模型预测

目的针对犯罪现场血液遗留时间的判断问题,实验以不同遗留时间的血液为研究对象,建立基于拉曼光谱的血液遗留时间的辨别方法。方法采集遗留时间为0.5h~240h的指尖血液样本,获取其拉曼光谱数据,经校正和平滑处理后,对数据进行归一化。利用RSD值评价光谱的稳定性,选取前10个主成分和6个重要波段分别建立模型,利用预测集测试模型效果。结果相同遗留时间血液的拉曼光谱具有较好的稳定(RSD<0.2),不同遗留时间血液的拉曼光谱在680cm^-1等6个波段有明显的强度变化。前10个主成分建立的PLSR模型,其r²=0.9927。6个重要波段建立的PLSR模型,预测集效果r²=0.8797。结论利用拉曼光谱结合建模分析是一种无损快速的评估血液遗留时间的方法。

Study and model prediction of blood remaining time based on Raman spectroscopy

Objective In order to judge blood remaining time at crime scene,the experiment takes blood of different remaining time as the research object,and establishes the identification method of blood remaining time based on Raman spectrum.Methods Fingertip blood samples with a residual time of 0.5 h~240 h were collected to obtain their Raman spectral data,which were normalized after correction and smoothing processing.RSD values were used to evaluate the stability of the spectrum,the top 10 principal components and six important bands were selected to build the model,and the model effect was tested by using the prediction set.Results The Raman spectra of blood with the same remaining time were stable(RSD<0.2).Raman spectrum of blood at different remaining time showed obvious intensity changes in six bands such as 680 cm^-1.The R² of the PLSR model established by the first 10 principal components was 0.9927.The predicted set effect R²=0.8797 for the PLSR model established in six important bands.Conclusion Raman spectroscopy combined with modeling analysis is a nondestructive and rapid method to evaluate the remaining time of blood.