Implementation of A Mobile Phone Based Atrial Fibrillation Prediction Platform

• Main Authors: Ming-Hsi Tu, 涂銘熙
• Other Authors: Ching-Hsiang Tseng
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/59721550417353801436

碩士 === 國立臺灣海洋大學 === 電機工程學系 === 101 === In Taiwan, cardiovascular diseases are top three cause of death. Atrial fibrillation is a cardiobascular disease that can be viewed as a more serious arrhythmia. The probability of occurrence of this disease increased with age, the percentage of age over 65 is 5%. Through recurrent self-testing the cardiac rhythm status, one can reduce the probability of damage caused by cardiovascular diseases. In this study, a mobile phone is combined with an electrocardiogram measurement circuit by using bluetooth wireless transmission to serve as a device for self-testing the cardiac rhythm status. The poincare plot is used in the thesis to analyze the heart rate variability(HRV). The result can be used to predict atrial fibrillation. We use the R-R interval data of "The PAF Prediction Challenge Database" from the MIT-BIH to predict paroxysmal atrial fibrillation (PAF) in our poincare plot analysis. Normal people and PAF patients can be distinguished by using the standard poincare plot analysis. A modified poincare plot analysis is also proposed in this thesis. We show that, for those undetected PAF patients under the standard poincare plot analysis, the proposed modified poincare plot analysis can be used to improve the detection rate.