Detekce srdečních abnormalit z EKG signálu pomocí neuronových sítí

Abstract

This thesis focuses on the analysis of cardiac abnormalities from electrocardiogram (ECG) signals using deep neural networks. The primary objective is to evaluate the effectiveness of various approaches for detecting and classifying cardiac pathologies based on publicly available datasets. The thesis includes the design and implementation of a software application developed in MATLAB, which processes and analyzes ECG signals and compares experimental results. The study compares the performance of different neural network models, including convolutional neural networks (CNNs), hybrid CNN-LSTM models, and transformers, with respect to accuracy, sensitivity, specificity, and F1 score. The experimental results provide insights into the capabilities of modern deep learning methods for diagnosing cardiac abnormalities and their applicability in clinical practice.