Related Experiment Video
Updated: Jan 11, 2026
08:20
Brain Source Imaging in Preclinical Rat Models of Focal Epilepsy using High-Resolution EEG Recordings
Published on: June 6, 2015
15.8K
Exploring brain lobe-specific insights in an explainable framework for EEG-based schizophrenia detection
Md Milon Hossain1, Md Nurul Ahad Tawhid1
1Institute of Information Technology, University of Dhaka, Dhaka, Bangladesh.
Biorxiv : the Preprint Server for Biology
|November 19, 2025
Summary
This study introduces a novel framework using electroencephalography (EEG) mel-spectrograms and Convolutional Neural Networks (CNNs) for accurate schizophrenia (ScZ) detection. The approach enhances diagnostic explainability by analyzing brain lobe activity, achieving high accuracy.
Area of Science:
- Neuroscience
- Medical Imaging
- Artificial Intelligence
Background:
- Schizophrenia (ScZ) poses a significant global health challenge, necessitating improved diagnostic tools.
- Current Electroencephalography (EEG) methods for ScZ detection lack specific brain lobe biomarkers and explainability.
- Integrating Explainable AI (XAI) is crucial for trustworthy medical diagnoses.
Purpose of the Study:
- To develop an explainable EEG-based framework for Schizophrenia (ScZ) detection using Convolutional Neural Networks (CNNs).
- To identify critical brain regions and biomarkers for ScZ diagnosis through lobe-specific analysis.
- To enhance the trustworthiness and clinical utility of EEG diagnostics for ScZ.
Main Methods:
- EEG signals were transformed into mel-spectrogram images using Short-Time Fourier Transform (STFT).
- A CNN model was employed for classification between ScZ patients and healthy controls (HC).
- Explainable AI techniques (LIME, SHAP, Grad-CAM) were integrated for result interpretability.
Main Results:
- The proposed framework achieved high classification accuracy (up to 99.82%) on two public EEG datasets.
- The frontal lobe demonstrated the highest diagnostic significance, followed by the temporal lobe.
- Occipital lobe analysis showed lower accuracy, indicating reduced relevance for ScZ diagnosis in this context.
Conclusions:
- EEG analysis combined with CNNs and mel-spectrograms offers a powerful, explainable method for ScZ detection.
- Brain lobe-specific analysis using this framework provides valuable insights for clinical guidance and diagnosis.
- The integration of XAI techniques significantly enhances the clinical applicability and trustworthiness of EEG-based ScZ diagnostics.
