【摘要】 目的建立计算机智能辅助诊断模型,降低主观因素影响,提高麻醉并发症的预测能力,有效减少麻醉风险。方法针对肿瘤手术麻醉风险评估,设计BP神经网络三层结构,包括输入层、隐藏层、输出层。收集350条临床患者数据,提取29个麻醉并发症的影响因素,对模型进行训练,并对麻醉并发症进行预测。结果该神经网络模型实现对10种常见麻醉并发症进行预测,平均准确率为89.8%。结论可以利用训练好的神经网络作为肿瘤手术麻醉风险评估的临床预测辅助诊断模型。更多还原

【Abstract】 Objective To improve the ability of prediction and prevention of anesthesia complications and reduce the influence of subjective factors in effectively reducing the incidence of anesthesia risk. Methods According to the risk assessment of anesthesia in tumor surgery, three layers of BP neural network structure were designed which included input layer, hidden layer and output layer. Data of 350 clinical patients were collected and 29 characteristic values were extracted. Then the model was trained, and the anesthesia complications were predicted. ResultsTen common anesthesia complications were predicted by the neural network with an average accuracy of 89.8%. Conclusion The trained neural network can be used as a clinical predictive diagnostic model for tumor surgical anesthesia risk assessment.更多还原