Optimized adaptive neuro-fuzzy inference system based on hybrid grey wolf-bat algorithm for schizophrenia recognition...

Area of Science:

• Neuroscience
• Artificial Intelligence
• Medical Diagnostics

Background:

• Schizophrenia is a chronic mental disorder impacting cognition, emotion, and behavior.
• Accurate and timely diagnosis of schizophrenia is crucial for patient quality of life.
• Existing diagnostic methods can be improved with advanced computational approaches.

Purpose of the Study:

• To develop an accurate computer-assisted system for schizophrenia prediction using multi-channel electroencephalogram (EEG) signals.
• To enhance the performance of an adaptive neuro-fuzzy inference system (ANFIS) through optimization.
• To evaluate the efficacy of a novel optimization algorithm for diagnostic classification.

Main Methods:

• Pre-processing of EEG signals using Butterworth band pass filter and wICA.
• Extraction of statistical, time-domain, frequency-domain, and spectral features.
• Feature selection using the ReliefF algorithm.
• Classification using ANFIS optimized by the Hybrid Grey Wolf-Bat Algorithm (HWBO).
• Validation through tenfold cross-validation.

Main Results:

• Achieved high classification accuracy (99.54% and 99.35%) on two EEG datasets.
• Demonstrated superior performance of HWBO-optimized ANFIS compared to traditional ANFIS and other optimization algorithms.
• HWBO-ANFIS showed high R² values and low RMSE, indicating efficient parameter optimization.
• Tenfold cross-validation yielded accuracies of 97.8% and 98.5%.

Conclusions:

• The proposed HWBO-optimized ANFIS method offers a highly accurate and efficient approach for computer-assisted schizophrenia diagnosis from EEG signals.
• The Hybrid Grey Wolf-Bat Algorithm significantly improves ANFIS performance for schizophrenia prediction.
• This computational approach holds promise for enhancing the diagnosis and management of schizophrenia.