Decoding Pain Dynamics: EEG Insights into Neural Responses and Classification Via RQA Analysis

Mahsa  Tavasoli; Zahra  Einalou; Reza  Akhondzadeh

doi:10.18502/fbt.v12i3.19182

Mahsa Tavasoli Department of Biomedical Engineering, North Tehran Branch, Islamic Azad University, Tehran, Iran
Zahra Einalou Department of Biomedical Engineering, North Tehran Branch, Islamic Azad University, Tehran, Iran
Reza Akhondzadeh Department of Anesthesiology, Pain Research Center, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran

DOI: https://doi.org/10.18502/fbt.v12i3.19182

Keywords: Cold Pressor Test; Electroencephalogram; Phasic Pain; Rough Neural Network; Recurrence Quantification Analysis; Electroencephalogram Dynamics; Pain Assessment.

Abstract

Purpose: Pain detection remains a challenging aspect of medical diagnosis, necessitating innovative approaches to address existing limitations. Current pain detection methods often lack precision and efficiency, prompting the exploration of alternative methodologies. This study focuses on investigating dynamic Electroencephalogram (EEG) patterns during pain states, aiming to fill gaps in the current understanding of pain detection mechanisms.

Materials and Methods: EEG recordings were conducted on a cohort comprising ten participants (5 men, and 5 women) who were free from drug usage and underlying ailments. The participants underwent EEG recording sessions during both resting and phasic pain states induced by immersing their left hand in ice-cold water. The EEG data were subjected to rigorous analysis using Recurrence Quantification Analysis (RQA). Additionally, a rough neural network classifier, with specific parameters tailored to the dataset characteristics, was employed for pain state classification.

Results: Our analysis revealed dynamic EEG features during phasic pain states, elucidated through RQA. Notably, the rough neural network demonstrated a high classification accuracy of 95.25% in distinguishing between pain and non-pain states. While specific numerical results such as p-values are not provided, the robust accuracy of the classification underscores the discernibility of cerebral responses during painful experiences.

Conclusion: This study contributes to the advancement of pain detection methodologies by introducing an innovative approach that leverages EEG analysis and neural network classification. While further investigation is warranted to validate these findings, they hold promise for enhancing pain assessment accuracy and ultimately improving patient care outcomes.