Detection of atrial fibrilation in long-term ECG

The bachelor’s thesis deals with the automatic detection of atrial fibrillations in the long-term ECG signals. First, it provides a description of the electrophysiology of the heart, the atrial fibrillation and the automatic methods of their detection. The first method, implemented in this work, is based upon the parameters that were calculated from the irregularities of the RR intervals. The second method uses the stationary wavelet transform and other parameters are computed after the signal transformation. The calculated parameters are subsequently statistically evaluated in the STATISTICA software. Parameters are assessed by the non-parametric Mann-Whitney test, which selects parameters that exhibit statistically significant differences between signals containing atrial fibrillation and sinus rhythm. At the end, the classification is performed by two approaches such as Support vector machine and k-Nearest Neighbours