A comprehensive approach to proactive performance assessment in safety-critical industries through EEG monitoring and...

Enhancing human performance is crucial in various industries for improved operational efficiency and safety, as even minor fluctuations can lead to severe consequences. The integration of electroencephalography (EEG) and advanced analysis methods have become tailor-made for understanding and optimizing cognitive processes to mitigate such errors and accidents. This article delves into the realm of cognitive assessment and its implications for the optimization of human performance to forge a tool for predicting cognitive capacities. The methodology relies on the collection of EEG data, with a specific focus on the activity in the prefrontal cortex, which serves as an index for attention and working memory status. Ten healthy adults participated in these experiments, undergoing EEG measurements, and standardized cognitive tests in controlled environments over 15 d. The data analysis involved preprocessing EEG signals, feature extraction, and modeling using machine learning techniques including k-nearest neighbor (KNN), decision trees, support vector machines, and artificial neural network (ANN) models. The findings unequivocally single out the decision tree model as the leading performer among the machine learning techniques scrutinized. It impressively attained a sensitivity of 94.25%, underscoring its precision in identifying individuals with robust attentional performance. The model's precision soaring at 84.97% and accuracy at 83.47% reinforce its ability to differentiate true positive cases with a minimal margin of false positives. However, the ANN model stands out as the best performer among memory models with an impressive accuracy of 83.90%. These findings add on the potential of EEG signals and machine learning for practical applications, emphasizing the value of eye blink patterns and neurophysiological data in predicting cognitive performance.