基于人工神经网络的儿科机械通气撤机系统研究

Research on the Paediatric Weaning System of Mechanical Ventilation Based on the Artificial Neural Network

机械通气已经成为治疗儿科各种危重症患者的重要手段,对于改善通气和氧合起了关键作用,从而为改善原发病治疗或急救赢得了时间。然而撤机时间的把握是临床上的难点,儿科的特点造就了更多的困难。近年来人工神经网络在放射的影像诊断模型的建立、消化道肿瘤检测诊断、流行病学等医疗领域的应用都有了很多进展,在成人的机械通气撤机上也有少量报道。本研究构建和训练一个人工神经网络,应用于判断重症患儿机械通气撤机时机。有望减少由于撤机时机不好造成的再次上机、并发感染。从而期望减少住院时间、医疗经费和其他医疗资源的消耗,减少患儿病痛,具有较大的社会效益和经济效益。

The ventilator has become an important means of treating a variety of critically ill pediatric patients. Mechanical ventilation to improve ventilation and oxygenation played a key role, so as to improve the treatment of the primary disease or emergency earned time. However, the doctor is difficult to grasp mechanical ventilation weaning time, the characteristics of pediatric made more difficult. In recent years, artificial neural network established the model of radiation in diagnostic imaging, tumor detection and diagnosis of gastrointestinal medical field, epidemiology and other applications have a lot of progress. There also have a few reports on weaning from mechanical ventilation in adults. This research build and train an artificial neural network used in PICU to determine the timing of weaning from mechanical ventilation. This improvement is expected to decrease due to the treatment regimen weaning caused by bad timing on the machine again, concurrent infection. Thus expected to reduce the length of hospital stay, medical expenses and other medical resource consumption, reduce pain in children, with great social and economic benefits.