Applied the MoDWT and STFT Layers to Classify the EEG of Schizophrenia Patients

The Nowadays, deep learning methods and patterns using brain activity recorded by electroencephalography (EEG) are valuable resources for making this diagnosis. In this research, a novel method to identify schizophrenia seizures from electroencephalogram (EEG) signals is presented, based on the maximal overlap discrete wavelet transform (MoDWT) and Short-time Fourier Transform (STFT). To carry out this research, we investigated MODWT’s potential to break down the signals into time–frequency sub-bands up to the fifth level. Furthermore, STFT has been investigated to illustrate the spectrogram from the sub-bands. Additionally, STFT and MoDWT layers were used in the research instead of the two traditional methods, which have been integrated into the convolutional neural network (CNN) model. The design aimed to give a feature extraction layer with the capacity to acquire the ability to handle different signal forms while also inheriting the signal’s extraction capabilities. The primary outcome of the research is that the suggested method could denoise EEG and extract the features with satisfactory accuracy.