作者: W J Tompkins , Y H Hu , V X Afonso , J L Urrusti
DOI:
关键词: QRS complex 、 Pattern recognition 、 Ecg signal 、 Signal processing 、 Multilayer perceptron 、 Artificial intelligence 、 Artificial neural network 、 Computer science 、 Background noise
摘要: The authors have investigated potential applications of artificial neural networks for electrocardiographic QRS detection and beat classification. For the task detection, used an adaptive multilayer perceptron structure to model nonlinear background noise so as enhance complex. This provided more reliable complexes even in a noisy environment. complex pattern classification, network was classifier distinguish between normal abnormal patterns, well classify 12 different morphologies. Preliminary results using MIT/BIH (Massachusetts Institute Technology/Beth Israel Hospital, Cambridge, MA) arrhythmia database are encouraging.