基于人工神经网络的病人麻醉深度判断方法

Research on Evaluating Depth of Anesthesia Based on Artificial Neural Network

在全身麻醉手术过程中,麻醉深度监测十分重要。为了获得更精确、更简单以及非侵入式的麻醉深度方法,本文利用心跳变化相似程度来分析自主神经活性指标,以专家评估的意识程度指标建立类神经网络回归模型。针对人工神经网络初始权重不稳定的问题,研究了人工神经网络集成法,接着通过灵敏度和特异度分析方法判断自主神经活性指标回归前与回归后的精准度差别。最后使用T检测进行统计分析,实验结果表明基于人工神经网络获得病人麻醉深度模型方法具有较好的鲁棒性。

In nowadays surgery,depth of anesthesia is indispensable. For more accuracy and simple way to evaluate depth of anesthesia,in this study,the similarity index（ SI） is calculated from heart rate variability（ HRV）based on RR-interval of electrocardiogram（ ECG） via expert assessment of conscious Level（ EACL）. Ensemble neural networks（ ENN） is used to solve problem which is ANN unstable due to random initial weight problem,then the sensitivity and specificity analysis method was used to determine the accuracy of the similarity index before and after regression. Finally,T-test was used for statistical analysis. The experimental results show that the method based on artificial neural network for patient anesthesia depth model has better robustness.