Early Detection of Patients with Mild Cognitive Impairment through EEG-SSVEP-based Machine Learning Model

Abstract

Mild cognitive impairment (MCI) is a transitional stage between normal aging and Alzheimer's disease (AD), which is a progressive and irreversible condition. Detecting MCI at an early stage is crucial as it offers the last opportunity to intervene and potentially prevent the progression to AD. Therefore, there is a need for research on identifying biomarkers that can effectively screen for MCI. Recent research suggests that declines in perception and action occur prior to neurodegenerative changes, highlighting the need for biomarkers that assess the ventral and dorsal streams responsible for processing this information. Among the various biomarkers, electroencephalography with steady-state visual evoked potential (EEG-SSVEP) has shown significant promise, providing real-Time, high-resolution monitoring of brain activities in response to flicker stimulation and offering a measure of cognitive decline linked to MCI. Despite this, studies examining the visual pathway for early MCI detection are limited. In this study, we conducted research involving 24 healthy controls and 25 MCI patients to develop and validate a machine learning model for the early detection of MCI based on EEG-SSVEP data. Initially, we extracted 166 EEG-SSVEP biomarkers through a comprehensive analysis, including lobe power ratio, lobe connectivity ratio, and band connectivity ratio to assess visual pathway's response to stimuli and neural activity across frequency bands. These biomarkers were specialized in quantifying cognitive decline within the visual pathway, specifically within the dorsal and ventral streams of the brain. By employing a biomarker selection method, we identified six key EEG-SSVEP biomarkers as the most relevant for distinguishing between healthy controls and MCI patients. Subsequently, these six biomarkers were utilized to train a support vector machine for early detection of MCI. The results showed an accuracy rate of 95.69%, a sensitivity of 92.28%, and a specificity of 95.58%. This study offers valuable insights into enhancing the early detection of MCI by leveraging EEG-SSVEP data and machine learning to assess cognitive decline within the dorsal and ventral streams of the brain.