CGP17Pat: Automated Schizophrenia Detection Based on a Cyclic Group of Prime Order Patterns Using EEG Signals

Abstract

Background and purpose: Machine learning models have been used to diagnose schizophrenia. The main purpose of this research is to introduce an effective schizophrenia hand-modeled classification method.

Method: A public electroencephalogram (EEG) signal data set was used in this work, and an automated schizophrenia detection model is presented using a cyclic group of prime order with a modulo 17 operator. Therefore, the presented feature extractor was named as the cyclic group of prime order pattern, CGP17Pat. Using the proposed CGP17Pat, a new multilevel feature extraction model is presented. To choose a highly distinctive feature, iterative neighborhood component analysis (INCA) was used, and these features were classified using k-nearest neighbors (kNN) with the 10-fold cross-validation and leave-one-subject-out (LOSO) validation techniques. Finally, iterative hard majority voting was employed in the last phase to obtain channel-wise results, and the general results were calculated.

Results: The presented CGP17Pat-based EEG classification model attained 99.91% accuracy employing 10-fold cross-validation and 84.33% accuracy using the LOSO strategy.

Conclusions: The findings and results depicted the high classification ability of the presented cryptologic pattern for the data set used.

Keywords: EEG classification; NCA; cyclic group of prime order pattern; kNN; machine learning; schizophrenia detection.

Figure 1

Schematic diagram of the proposed CGP17Pat-based schizophrenia detection model.

Figure 2

The created eight patterns using CGP17, where each pattern is named as P; by using each pattern, 256 features are extracted and our presented CGP17Pat uses these eight patterns together: (a) patterns 1–4; (b) patterns 5–8.

Figure 2

The created eight patterns using CGP17, where each pattern is named as P; by using each pattern, 256 features are extracted and our presented CGP17Pat uses these eight patterns together: (a) patterns 1–4; (b) patterns 5–8.

Figure 3

The graphical summarization of the presented CGP17Pat. Here, P denotes patterns (see Table 2), and each pattern extracts 256 features. Then, these feature vectors are merged, and 2048 (=256 × 8) features are created.

Figure 4

The lengths of the optimal feature vectors chosen by INCA.

Figure 5

Classification accuracies of the decision tree (DT), quadratic discriminant (QD), logistic regression (LR), naive Bayes (NB), support vector machine (SVM), Fine kNN (kNN), bagged tree (BT), ensemble subspace kNN (ESkNN), and artificial neural network (ANN) for the Fp2 channel with 10-fold cross-validation.

Figure 6

The obtained comparative results according to the validation technique.

Figure 7

Voted results: confusion matrices of the presented CGP17Pat-based EEG classification model using (a) 10-fold cross-validation and (b) LOSO cross-validation.