【摘要】 提出一种基于人工神经网络的、利用多种描述信息检测心室纤颤的综合系统：首先，利用不同方法从ECG信号中提取出多种具有不同意义的描述信息，然后用神经网络的综合概括能力把它们融合进检测系统并自动作出决策。对该系统的检测性能进行了测试，最后，讨论了临床应用的可行性。更多还原

【Abstract】 A novel comprehensive system based on artificial neural networks (ANN) for detecting ventricular fibrillation (VF) using various signal descriptors is proposed. First, by using time-, frequency-domain, and nonlinear dynamics analyses, different kinds of descriptors with different meanings are extracted from cardiac rhythmic signals, and then incorporated into the detection system and make decision automatically via the comprehensive ability of ANN. The detection performance of the system is tested by actual ECG database. Finally, its feasibility in clinical application is discussed.更多还原