基于样本熵算法的早产儿败血症发病预测模型探索

Exploration on Predictive Model of Neonatal Sepsis Based on Sample Entropy Algorithm

针对早产儿患败血症几率较高但传统诊断较为滞后,易造成永久神经损伤的问题,提出了基于样本熵算法的早产儿败血症发病预测模型。将样本熵算法应用于心率变异分析,并根据心率数据特点优化选择了算法参数;然后,分析了基于本算法得出的样本熵变化与新生儿败血症发病的关联性和预测作用。结果表明,模型对于早产儿败血症发病预测结果与实际病程相符,对进一步研究提供了方向。

This article focuses on the prediction model of the incidence of sepsis in preterm children based on Sample Entropy algorithm, optimizes the chosen algorithm parameters, and analyses the relevance between the predict tive algorithm＇ s results and the neonatal sepsis onset.