An automated alcoholism detection using orthogonal wavelet filter bank

Abstract

Alcohol misuse is a common social issue related to the central nervous system. Electroencephalogram (EEG) signals are used to depict electrical activities of the brain. In the proposed study, a new computer-aided diagnosis (CAD) has been developed to recognize alcoholic and normal EEG patterns, accurately. In this paper, we present an automatic system for the classification of normal and alcoholic EEG signals using orthogonal wavelet filter bank (OWFB). First, we derive sub-bands (SBs) of EEG signals. Then, we compute logarithms of the energies (LEs) of the SBs. The LEs are employed as the discriminating features for the separation of alcoholic and normal EEG signals. A supervised machine learning algorithm called K nearest neighbor (KNN) has been employed to classify normal and alcoholic patterns. The proposed model has yielded very good classification results. We have achieved a classification accuracy (CA) of 94.20% with tenfold cross-validation (CV).