Baumert, MathiasAbbott, DerekEbrahimpour, Maryam2025-06-132025-06-132024https://hdl.handle.net/2440/145187The electroencephalogram (EEG) stands out as a valuable non-invasive tool for monitoring brain activities. However, EEG signals are prone to artefacts, including those originating from cardiac activity, which complicates accurate data interpretation. This research explores techniques specifically designed to eliminate cardiac artefacts from EEG signals, essential for enhancing the accuracy of biomedical signal processing. Two methods of cardiac artefact removal are developed and investigated in this thesis. The first method involves rejecting excessively contaminated segments and can be applied online and in real-time to single-channel EEG when an ECG signal is available. The second study entails evaluating four common ICA methods to specifically extract cardiac artefact component from 16-channel EEG. The investigation concludes that second order blind identification (SOBI) is the most effective algorithm for this specific application. Moreover, this thesis probes EEG potential as a biomarker, focusing particularly on heartbeat-evoked potentials (HEP) and frontal asymmetry. In relation to HEP, which is a crucial measure of the connectivity between the heart and the brain, our findings indicate that HEP is most prominent in fronto-central areas and the frontal lobe, supporting the hypothesis that considers the insula and somatosensory cortex as the source of HEP in the brain. Additionally, HEP reaches its maximum value during the REM stage of sleep, suggesting that the heightened vigilance during REM sleep may affect HEP strength. However, as a biomarker for recovery from sleep apnea and its related symptoms, we find that HEP does not effectively differentiate between surgery and control group. Another study performed in this thesis focuses on frontal asymmetry as a biomarker for depression in children using sleep EEG. Our findings do not demonstrate a significant relationship between frontal asymmetry and depression. Collectively, this thesis contributes to the advancement of EEG signal processing, artefact removal methodologies, and the utilization of physiological markers in clinical research.enElectroencephalogram (EEG)EEG artefactsbiomedical signal processingcardiac artefact removalIndependent Component Analysis (ICA)Second Order Blind Identification (SOBI)heartbeat-evoked potentials (HEP)frontal asymmetrysleep apneadepressionThesis