Automatizované hodnocení kvality transabdominálního fetálního EKG pro domácí monitorovací zařízení

Abstract

This diploma thesis focuses on the automated assessment of transabdominal fetal electrocardiogram (ECG) signal quality, which represents a promising non-invasive method for long-term monitoring of fetal cardiac activity. The quality of fetal ECG signals is significantly influenced by biological, anatomical, and technical factors, which complicates their reliable analysis. The aim of this work was to design and validate a system for automatic classification of fetal ECG signal quality using machine learning methods. After extracting time-domain and frequency-domain features, both supervised and unsupervised learning methods were tested. The best results were achieved using the Random Forest algorithm. The findings show that the combination of well-selected features and machine learning techniques enables effective and objective signal quality evaluation, paving the way for broader application of this technology in home and clinical monitoring systems.