A novel approach for automated alcoholism detection using Fourier decomposition method

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Alcoholism poses significant risks to the central nervous system and overall health, leading to conditions like cardiomyopathy, immune disorders, and cirrhosis.
• Early and accurate identification of alcoholism is crucial for mitigating its severe health consequences.
• Electroencephalogram (EEG) signals offer a potential biomarker for neurological changes associated with alcoholism.

Purpose of the Study:

• To develop and validate a novel, automated method for alcoholism identification using EEG signals.
• To leverage Fourier theory for enhanced signal decomposition and feature extraction from EEG data.
• To assess the efficacy of the proposed method in accurately classifying alcoholism.

Main Methods:

• A Fourier decomposition method (FDM) was employed to decompose EEG signals into Fourier intrinsic band functions (FIBFs).
• Time-domain features were extracted from FIBFs, and the Kruskal-Wallis test was used for feature selection.
• Classification was performed using k-nearest neighbor (kNN), support vector machine (SVM), and linear discriminant analysis (LDA).

Main Results:

• The proposed FDM approach, particularly with an SVM classifier, achieved exceptional accuracy (99.98%), sensitivity (99.99%), and specificity (99.97%).
• The method demonstrated robust performance even in the presence of noise.
• Results significantly outperformed existing methods for alcoholism detection.

Conclusions:

• The Fourier decomposition method (FDM) provides a highly accurate and efficient approach for automatic alcoholism identification.
• The proposed scheme is suitable for real-time applications in clinical settings.
• This novel method holds promise for improving the early detection and management of alcoholism.