Improving EEG Classification of Alcoholic and Control Subjects Using DWT-CNN-BiGRU with Various Noise Filtering Techniques

Nidhi Hitendrabhai Patel^1,2*Jaiprakash Verma^1*Swati Jain¹

• ¹Institute of Technology, Nirma University, Ahmedabad, India
• ²Nirma University, Ahmedabad, Gujarat, India

EEG(Electroencephalogram) signal analysis plays a vital role in diagnosing and monitoring alcoholism, whereaccurate classification of individuals into alcoholic and control groups is essential. However, the inherent noise and complexity of EEG signals pose significant challenges. This study investigates the impact of three signal denoising techniques—Discrete Wavelet Transform(DWT), Discrete Fourier Transform(DFT) and DiscreteCosine Transform(DCT) on the classification performance of EEG signals. The motivation behind this workis to identify the most effective preprocessing method for enhancing deep learning model performance inthis domain. A novel DWT-CNN-BiGRU model is proposed, which leverages CNN layers for spatial fea-ture extraction and BiGRU layers for capturing temporal dependencies. Experimental results show that theDWT-based approach, combined with standard scaling, achieves the highest accuracy of 94%, with a preci-sion of 0.94, recall of 0.95, and F1-score of 0.94. Compared to the baseline DWT-CNN-BiLSTM model, theproposed method provides a modest yet meaningful improvement of approximately 17% in classification ac-curacy. These findings highlight the superiority of DWT as a preprocessing method and validate the proposedmodel’s effectiveness for EEG-based classification, contributing to the development of more reliable medicaldiagnostic tools.