Table 1 List of published works on schizophrenia classification using EEG signals in recent years.
From: A hybrid deep neural network for classification of schizophrenia using EEG Data
Author (year) | EEG dataset | Rest/task | Sampling rate | Channels | Features | Classifier | Accuracy |
|---|---|---|---|---|---|---|---|
Bose et al., 201616 | 57 schizophrenia patients and 24 normal subjects | Rest | 256 | 23 | Absolute power analysis | SVM | 83.33% |
Johannesen et al., 201653 | 40 schizophrenia patients and 12 healthy controls | Task | 1024 | 60 | Morlet continuous wavelet transform | SVM | 87% |
Jeong et al., 201754 | 30 schizophrenia patients and 15 controls | Task | 1024 | 14 | Mean subsampling technique | SKLDA | Over 98% |
Piryatinska et al., 201755 | 45 boys suffering from schizophrenia and 39 healthy boys | Rest | 128 | 16 | є-complexity of a continuous vector function | RF | 85.3% |
Chu et al., 201756 | 10 normal and 17 markedly ill schizophrenic patients | Task | 256 | 31 | ApEn | SVM | 81.5% |
Alimardani et al., 201857 | 26 subjects with schizophrenia and 27 patients with BMD | Rest | 250 | 22 | DB-FFR | NN | 87.51% |
Alimardani et al., 2018 58 | 23 bipolar disorder and 23 schizophrenia subjects | Rest | 250 | 21 | SSVEP SNR | KNN | 91.30% |
Phang et al., 201959 | 45 schizophrenia patients and 39 healthy controls | Rest | 128 | 16 | Vector-autoregression-based directed connectivity (DC), graph-theoretical complex network (CN) | DNN-DBN | 95% |
Phang et al.,201960 | 45 schizophrenia patients and 39 healthy controls | Rest | 128 | 16 | Directed connectivity measures (VAR coefficients and PDCs) and topological CN measures | MDC-CNN | 91.69% |
Oh et al.,201914 | 14 healthy subjects and 14 SZ patients | Rest | 250 | 19 | – | CNN | 98.07% |
Present work | 54 patients with schizophrenia and 55 healthy controls | Rest | 500 | 60 | FuzzyEn | CNN + LSTM | 99.22% |
