NON-LINEAR DYNAMICS IN BRAIN ACTIVITY DURING MENTAL ARITHMETIC TASKS

Abstract

Mental arithmetic tasks are cognitively demanding activities that require sustained attention, working memory, and numerical reasoning. These processes are associated with dynamic and distributed patterns of brain activity, which can be captured through electroencephalographic (EEG) signals. In this work, a comprehensive analysis was conducted to investigate the non-linear dynamics of brain activity before and during mental arithmetic, aiming to capture the differences between both states.

EEG signals were acquired from healthy university students during both resting state and mental arithmetic conditions. The analysis focused on electrode O2, given its relevance to visual and cognitive processing. To characterize brain dynamics, a combination of linear methods, power spectral analysis, and non-linear techniques, Sample Entropy and Correlation Dimension, were employed to extract features sensitive to changes in brain activity.

Spectral analysis revealed an increase in power within frequency bands associated with cognitive activation during arithmetic tasks, alongside a decrease in frequencies typically linked to rest and relaxation. Non-linear measures showed a significant reduction in Sample Entropy and Correlation Dimension during task performance, indicating a transition toward more organized and predictable neural dynamics under cognitive load.

In addition, the study explored the influence of gender and performance on EEG dynamics. While cognitive load induced clear changes in brain activity, no statistically significant differences were observed based on gender or task performance levels.

This work demonstrates the potential of non-linear EEG analysis for detecting cognitive load related changes and highlights the value of combining multiple analytical approaches. The findings lay the groundwork for future applications in educational and clinical settings, such as the development of tools for real-time assessment of mental effort.