| Abstract [eng] |
Obstructive sleep apnea (OSA) is a common sleep disorder marked by recurring constraints of the upper airway, which cause reduced or halted breathing during sleep. Globally, over a billion individuals are affected by sleep apnea syndrome, with some countries reporting prevalence rates above 50%. OSA patients often require increased healthcare services and have a higher likelihood of developing type 2 diabetes, mood and anxiety disorders, and cardiovascular diseases. In addition, latest research suggests potential associations between OSA and cardiac arrhythmias such as atrial fibrillation. In medical practice, the apnea-hypopnea index (AHI) serves as the benchmark for identifying the presence and severity of OSA. However, the AHI has been criticized for failing to capture pertinent clinical characteristics and being an inadequate tool for predicting clinical outcomes. At present, the significance of the AHI as a diagnostic metric of severity for clinically relevant OSA is diminishing. This thesis argues that the AHI alone is insufficient for evaluating OSA severity, as it only measures the number of events per hour of sleep and does not account for other crucial factors such as potential cardiovascular effects. In order to monitor relevant cardiovascular effects in OSA patients, electrocardiogram (ECG) and photoplethysmogram (PPG) signal analysis can be utilized. This doctoral thesis proposes and explores a novel cross-recurrence properties-based method for characterizing OSA severity, considering potential association to cardiac arrhythmias. The proposed method estimates cross-recurrence properties between specific feature time series, extracted from ECG and PPG signals. |
