Akademik Veri Yönetim Sistemi
33rd Conference on Signal Processing and Communications Applications-SIU-Annual, İstanbul, Türkiye, 25 - 28 Haziran 2025, (Tam Metin Bildiri)
In this study, the Tuneable Q-factor Wavelet Transform (TQWT) method was applied to classify Attention Deficit and Hyperactivity Disorder (ADHD) diagnosis with features obtained from EEG signals. It was aimed to perform EEG-based ADHD diagnosis more objectively and reliably. EEG signals were analyzed in the prefrontal, frontal, central, parietal, temporal, and occipital regions. The importance levels of the features obtained from the four sub- bands of TQWT and the signal are determined by ANOVA test and classified by machine learning algorithms. In the evaluations made with 10-fold cross-validation, 99.34% training, and 99.41% test accuracy were obtained in the parietal region in the Ensemble Subspace k-NN model, 98.90% training accuracy and 98.53% test accuracy were obtained in the frontal region. The results show that the TQWT method successfully detects important patterns in EEG signals and reveals the cognitive and neurophysiological differences of ADHD.