Machine-Learning-Based Elderly Stroke Monitoring System Using Electroencephalography Vital Signals

Stroke is the third highest cause of death worldwide after cancer and heart disease, number stroke diseases due to aging set at least triple by 2030. As top three causes are all related chronic importance healthcare increasing even more. Models that can predict real-time health conditions using various services attracting attention. Most diagnosis prediction methods for elderly involve imaging techniques such as magnetic resonance (MRI). It difficult rapidly accurately diagnose long testing times high costs associated with MRI. Thus, in this paper, we design implement a monitoring system precursors real time during daily walking. First, raw electroencephalography (EEG) data from six channels were preprocessed via Fast Fourier Transform (FFT). The EEG power values then extracted spectra: alpha (α), beta (β), gamma (γ), delta (δ), theta (θ) well low β, θ β ratio, respectively. experiments paper confirm important features biometric signals alone walking determine occurrence more than 90% accuracy. Further, Random Forest algorithm quartiles Z-score normalization validates clinical significance performance proposed 92.51% be implemented cost, it applied early disease detection precursor symptoms stroke. Furthermore, expected will able detect other future.