In this study, we investigate the effects of Adversarial Neural Network Training (ANNT) on the robustness and effectiveness of Brain-to-Brain Communication (B2B-C) systems using Steady-State Visually Evoked Potentials (SSVEP) EEG data. We utilized a combined Convolutional Neural Network-Temporal Convolutional Network (CNN-TCN) architecture to classify the data and assessed the system's resistance to various adversarial strategies, including Fast Gradient Sign Method (FGSM), Projected Gradient Descent (PGD), Basic Iterative Method (BIM), Carlini & Wagner (C&W), and Momentum Iterative Method (MIM). By analyzing publicly accessible datasets, specifically Lee2019_SSVEP and Nakanishi2015, we observed significant enhancements in both accuracy and AUC metrics when ANNT was applied. Specifically, the Lee2019_SSVEP dataset exhibited a 24% increase in accuracy and a 0.23-point improvement in AUC, while the Nakanishi2015 dataset demonstrated improvements of 9% and 0.07 points, respectively. Our results indicate that PGD posed the greatest challenge to the model, significantly reducing accuracy and AUC across various scenarios, whereas FGSM was the least impactful. These findings highlight ANNT's potential in fortifying the security and stability of B2B-C systems against diverse adversarial conditions.

This study proposes an innovative hierarchical binary classification approach based on scalp EEG data to predict comatose patients' Glasgow Outcome Scale (GOS).The dataset consists of continuous EEG recordings from 13 comatose patients classified into three GOS groups: GOS 1 (death), GOS 3 (severe disability), and GOS 5 (good recovery). Traditional machine learning (ML) and deep learning (DL) models, though effective in some instances, struggle to generalize across diverse patient populations, mainly when evaluated with leave-one-subject-out (LOSO) cross-validation. We introduce a two-step binary classification approach to address these challenges, utilizing Convolutional Neural Networks (CNNs) with spectrogram features for one classifier and Random Forest (RF) with Approximate Entropy (ApEn) for the other. By simplifying the multi-class classification task into two binary decisions, we achieve perfect classification performance for GOS 1 and GOS 5, while GOS 3 is inferred by exclusion. Importantly, we found that just 2 minutes of EEG data per subject was sufficient to achieve this level of performance, offering significant implications for clinical applications where data is often limited. Even considering the small dataset, our results prove that our method performs better than traditional approaches, providing a scalable, efficient solution for coma outcome prediction.