Hybrid Machine Learning for Cardiovascular Risk Assessment in Obstructive Sleep Apnea Patients

The study looks at how Obstructive Sleep Apnea (OSA) affects cardiovascular disease (CVD). It details the ways in which OSA increases the risk of cardiovascular problems involving hypoxia, oxygen depletion, high blood pressure, and a host of others. The SHHS information was utilized to create a hybrid approach to learning that can categorize individuals with OSA into three hazard groups: On the cardiovascular disease (CVD) risk scale, including 0–5%, 10%, and 20%. The model achieved a classification accuracy of 98.89% in high-risk patients and 97.87% in no-risk patients using the ResNet101 DNN combined with decision trees. Additionally, the study takes into account HDL saturated fat, blood pressure, and cholesterol in connection to the risk of CVD. More scholarships have shown that increased oxidative stress is a major risk influence for CVD in OSA patients. Sleep and heart rate, two crucial metrics for risk categorization, were measured using a polysomnography (PSG). By highlighting the need for a real-time assessment of CVD risk in OSA patients, the present suggested model outperforms earlier methods. Further studies will expand it by incorporating other factors like stress and obesity to enhance the machine learning model's prediction capabilities.