Arrthymia Detection through the Combination of Feature Extraction and Hidden Markov Models

• Main Authors: Chin-Lun Kuo, 郭金倫
• Other Authors: 林志麟
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/01058234184597859282

碩士 === 元智大學 === 資訊管理學系 === 98 === This study proposes a method to detect the occurrence of Arrhythmia in ECG signals. Arrhythmia means abnormal cardiac rhythms that contain abnormal rhythms. It may cause patient indisposed, or even died. So the detection of abnormal cardiac rhythms from normal heart activity became more important. First we extract and then discretize feature data from original ECG, then discriminative, Then, we create an Arrhythmia models array from Hidden Markov Models. In the testing step, the ECG data will be segmented by a sliding window. We will get a probability array by taking every segment into Arrhythmia models array. Finally we calculate the probability of Arrhythmia to each point in the ECG signal.