Cloud EEG Privacy Using Red-Billed Blue Magpie Optimized Physics-Penalized Dual-Branch Spectral-Spatial Neural...

Area of Science:

• Neuroscience and Biomedical Engineering
• Artificial Intelligence in Healthcare

Background:

• Epileptic seizure prediction is crucial for timely intervention and preventing complications.
• Real-time electroencephalogram (EEG) monitoring via the Internet of Things (IoT) is vital for automated seizure detection.

Purpose of the Study:

• To develop a hybrid deep learning architecture for precise, safe, and effective epileptic seizure prediction.
• To enhance the accuracy and reliability of seizure prediction using advanced signal processing and machine learning techniques.

Main Methods:

• Real-time EEG data acquisition using an IoT-based headband.
• Signal denoising and enhancement with Shape-Aware Mesh Normal Filtering (SMNF).
• Feature extraction using Quadratic Phase Quaternion Domain Fourier Transform (QPQDFT).
• Classification via Physics-Penalized Dual-Branch Spectral-Spatial Neural Network (PP-DBSSNN) with physics-based regularization and dual-branch attention.
• Secure data handling using Key Escrow-Free Attribute-Based Encryption (KEF-ABE).

Main Results:

• Achieved 99.95% accuracy, 99.93% precision, and 99.91% specificity on the Bonn EEG dataset.
• Achieved 99.96% accuracy, 99.94% precision, and 99.92% specificity on the CHB-MIT dataset.
• Demonstrated robustness and reliability of the proposed hybrid deep learning approach.

Conclusions:

• The hybrid deep learning architecture offers a highly accurate and reliable solution for real-time epileptic seizure prediction.
• The integration of advanced signal processing, deep learning, and secure encryption methods enhances EEG data analysis for clinical applications.
• This approach holds significant potential for improving patient outcomes and managing epilepsy effectively.