Classification of obstructive sleep apnea using bio-control feedback EEG biosensor device

Obstructive sleep apnea (OSA) is a common sleeping disorder that can result in shortness of breath due to the relaxation of upper respiratory tract muscles while the patient is asleep. This leads to low blood oxygen saturation and, hence, apnea. OSA is associated with higher risks of hypertension, coronary artery disease, arrhythmias, cardiac failure, and stroke, affecting 5% of the global human population. Thus, early detection and treatment of OSA are crucial. Nowadays, OSA is diagnosed by analyzing EEG signals. Electroencephalographic (EEG) signals are important for monitoring brain activity; this paper focuses on patients with OSA. Unfavorable spikes in the graph, such as head movements, electrical interferences, or even muscle spasms, may interfere with the general EEG frequencies being targeted. Our research was centered on designing a cost-effective electronic circuitry for this device that would efficiently acquire and process EEG signals. We used various methods to improve the device's functionality and accuracy, including rectifications in noise reduction, signal filtering, and structural correction through ICA decomposition. ⁤⁤Our system, theoretically, achieved a high accuracy rate of 97.14% using Support Vector Machines.