Optimizing motor imagery BCI models with hard trials removal and model refinement.

Deep learning models have demonstrated remarkable performance in the classification of motor imagery BCI systems. However, these models exhibit sensitivity to challenging trials, often called hard trials, leading to performance degradation. In this paper, we address this issue by proposing two novel methods for identifying and mitigating the impact of hard trials on model performance. The first method leverages model prediction scores to discern hard trials. The second approach employs a quantitative explainable artificial intelligence (XAI) approach, enabling a more transparent and interpretable means of hard trial identification. The identified hard trials are removed from the entire motor imagery training and validation dataset, and the deep learning model is further re-trained using the dataset without hard trials. To evaluate the efficacy of these proposed methods, experiments were conducted on the Open BMI dataset. The results for hold-out analysis show that the proposed quantitative XAI- based hard trial removal method has statistically improved the average classification accuracy of the baseline deep CNN model from 63.77% to 68.70%, withp-value =7.66e-11for the subject-specific MI classification. Additionally, analyzing the scalp map representing the average relevance scores of correctly classified trials compared to a misclassified trial provides a deeper insight into identifying hard trials. The results indicate that the proposed quantitaive-based XAI approach outperformes the prediction-score-based approach in hard trial identification.

Improved Patient-Independent System for Detection of Electrical Onset of Seizures.

PURPOSE: To design a non-patient-specific system to detect the electrical onset of seizures in patients with temporal lobe epilepsy. METHODS: We used EEG data from 29 seizures of 18 temporal lobe epilepsy patients who underwent multiday video-scalp EEG monitoring as part of their presurgical evaluations. We segmented each data set into preictal and ictal phases, and identified spectral entropy, spectral energy, and signal energy as useful features for discriminating normal and seizure conditions. The performance of five different classifiers was analyzed using these features to design an automated detection system. RESULTS: Among the five classifiers, decision tree, k-nearest neighbor, and support vector machine performed with sensitivity (specificity) of 79% (81%), 75% (85%), and 80% (86%), respectively. The other two, linear discriminant algorithm and Naive Bayes classifiers, performed with sensitivity (specificity) of 54% (94%), 47% (96%), respectively. CONCLUSIONS: The support vector machine-based seizure detection system showed better detection capability in terms of sensitivity and specificity measures as compared to linear discriminant algorithm, Naive Bayes, decision tree, and k-nearest neighbor classifiers. CONCLUSIONS: Our study shows that a generalized system to detect the electrical onset of seizures in temporal lobe epilepsy using scalp-recorded EEG is possible. If confirmed on a larger data set, our findings may have significant implications for the management of seizures, especially in patients with drug-resistant epilepsy.