Alpha Anterior Posterior Index as a Novel Quantitative EEG Biomarker for Alzheimer's Disease

Alzheimer's disease (AD) is the most prevalent form of dementia. Previous studies showed some promising biomarkers using quantitative electroencephalograms (qEEG), but more are waiting to be explored. This study aims to optimize and identify novel biomarkers to aid in diagnosing AD. This study uses a publicly available preprocessed EEG recording dataset from 19 electrodes to investigate power spectral density (PSD) patterns in AD and control (CN) groups. A novel PSD-derived biomarker, Alpha Anterior Posterior Index (aAPI), along with previously reported Alpha Relative PSD, Theta Alpha Ratio (TAR), and Spectrum Ratio (SR) biomarkers, are studied in diagnosing AD. Statistical analyses and machine learning using logistic regression models are employed to test the performance of these biomarkers. The results of our analysis reveal significant differences between AD and CN in the Alpha Relative PSD, Theta Relative PSD, TAR, SR, and the novel biomarker aAPI, with the Alpha band showing decreased power, Theta band showing increased power, TAR being higher, SR being lower, and aAPI being lower in AD compared to CN. Machine learning models, particularly the combination of aAPI and TAR, prove most effective in distinguishing between AD and CN, achieving the highest accuracy among all biomarkers. Individual biomarkers such as Alpha Relative PSD and Spectrum Ratio also exhibit relatively high accuracy, albeit not surpassing the combination of aAPI and TAR. The findings of this study suggest promising results for aAPI as a novel biomarker.