TABLE 2.
Studies using Deep Learning methods for detection of SCZ.
| Study | Pre-processing | Method | Signal | Database | Accuracy (%) |
| Bagherzadeh et al., 2022 | Butterworth filters and Transfer Entropy (TE) | VGG-16, ResNet50V2, InceptionV3, EfficientNetB0, DenseNet121, and CNN-LSTM | EEG | IPN (Olejarczyk and Jernajczyk, 2017) | 99.90 |
| Aslan and Akin, 2022 | Continuous Wavelet Transform | Visual Geometry Group-16 (VGG16), an Advanced convolutional neural network (CNN) | EEG | 1. Private 2. IPN (Olejarczyk and Jernajczyk, 2017) |
99.5 |
| Supakar et al., 2022 | Dimensionality reduction algorithm | RNN-LSTM | EEG | Laboratory for Neurophysiology and Neuro-Computer Interfaces (Shishkin et al., 2011) | 98 |
| Shalbaf et al., 2020 | Wavelet Transform | AlexNet, ResNet-18, VGG-19, Inception-v3, and SVM | EEG | IPN (Olejarczyk and Jernajczyk, 2017) | 98.60 |
| Soria Bretones et al., 2023 | Digital filters and Fuzzy means clustering | Artificial neural network based on RBF | EEG | Psychiatry Department of Virgen de la Luz Hospital in Cuenca | 93 |
| (Siuly et al., 2023) | Average filtering | Deep ResNets, softmax layer and deep features with SVM | EEG | Kaggle SCZ dataset (Ford et al., 2014) | 99.23 |
| Siuly et al., 2022 | Average filtering | GoogleNet and deep features, SVM | EEG | Kaggle SCZ dataset (Ford et al., 2014) | 98.84 |
| Shoeibi et al., 2021 | Segmentation, Denoising, Normalization | 1-D-CNN, LSTM, and 1-D-CNN + LSTM | EEG | IPN (Olejarczyk and Jernajczyk, 2017) | 99.25 |
| Phang et al., 2019 | Connectivity measures (VAR coefficients and PDCs) | Multi domain-CNN | EEG | EEG SCZ (Gorbachevskaya and Borisov, 2002) | 91.69 |
| Singh et al., 2021 | Bandpass filter, segmentation, and Fast Fourier transform (FFT) | CNN + LSTM | EEG | 1. EEG SCZ (Gorbachevskaya and Borisov, 2002) 2. IPN (Olejarczyk and Jernajczyk, 2017) |
98.56 |
| Khare et al., 2021 | Short Term FFT, Continuous WT, and SPWVD | AlexNet, ResNet50, VGG16, and CNN | EEG | Kaggle SCZ dataset (Ford et al., 2014) | 93.36 |
| (Sun et al., 2021) | Fourier transform (FT) | CNN + LSTM | EEG | Huilongguan Hospital | 99.22 |
| Supakar et al., 2022 | Random projection | RNN + LSTM | EEG | Laboratory for Neurophysiology and Neuro-Computer Interfaces (Shishkin et al., 2011) | 98 |
| Hassan et al., 2023 | Digital bandpass filter | Hybrid Classifier (CNN + ML) | EEG | IPN (Olejarczyk and Jernajczyk, 2017) | 98.05 |
| Bao et al., 2023 | Notch filter + Principal component analysis (PCA) | CNN and temporal convolution networks (TCNs) | EEG | IPN (Olejarczyk and Jernajczyk, 2017) | 99.57 |
| Shen et al., 2023 | Continuous wavelet transform (CWT) + Digital bandpass filter | Customized 3D-CNN | EEG | Laboratory for Neurophysiology and Neuro-Computer Interfaces (Shishkin et al., 2011) | 98.89 |