Enhanced schizophrenia detection using multichannel EEG and CAOA-RST-based feature selection.

Area of Science:

• Neuroscience
• Medical Informatics
• Artificial Intelligence

Background:

• Schizophrenia diagnosis is challenging due to complexity and current technique limitations.
• Electroencephalogram (EEG) signals offer insights into brain activity but present high dimensionality and complexity issues.
• Accurate and early detection of schizophrenia is crucial for effective patient management.

Purpose of the Study:

• To develop and validate a novel, accurate, and efficient model for schizophrenia detection using EEG signals.
• To address the challenges of high-dimensional EEG data through advanced computational techniques.
• To improve upon existing state-of-the-art methods for schizophrenia diagnosis.

Main Methods:

• A four-stage model integrating multichannel EEG, Crossover-Boosted Archimedes Optimization Algorithm (CAOA), and Rough Set Theory (RST).
• Stages include data preprocessing (artifact removal, filtering, normalization), feature extraction (CAOA for selection, MEMF, entropy measures), and classification (Support Vector Machine - SVM).
• Model validated using real-world EEG datasets.

Main Results:

• The proposed model achieved high performance metrics: 94.9% average accuracy, 93.9% sensitivity, 96.4% specificity, and 92.7% precision.
• Demonstrated significant improvements over existing state-of-the-art methods.
• The integration of CAOA and RST effectively handled high-dimensional EEG data and optimized feature selection.

Conclusions:

• The novel EEG-based model shows significant potential for accurate and early schizophrenia detection.
• The combined approach of CAOA and RST offers a robust solution for analyzing complex, high-dimensional neurophysiological data.
• Future work should focus on larger, diverse datasets and advanced machine learning models to further refine diagnostic capabilities.