EEG and fNIRS Data for Pain Monitoring and Detection in a Neurofeedback System

Pain represents a disagreeable sensory and emotional sensation usually stemming from real or potential harm to bodily tissues [1]. It's a unique experience that varies from person to person due to its inherently subjective nature, influenced by biological, psychological, and social factors [2]. Over the course of life, individuals develop their understanding of pain. When it comes to conveying pain, it's not confined to verbal descriptions alone, as the inability to communicate doesn't negate the likelihood of a human or non-human animal feeling pain. Prior research has suggested investigating EEG and fNIRS signals in the context of pain detection triggered by painful stimuli [3], [4]. This study aims to analyze EEG and fNIRS signals in the frequency domain to detect pain, and it seeks to apply artificial intelligence to monitor data from portable EEG and fNIRS devices with the goal of developing a neurofeedback system. To conduct the measurements and develop the neurofeedback system, we have chosen a portable EEG helmet equipped with dry electrodes (CGX Quick-32r) and a Bluetooth fNIRS device (NIRSport2). In this study, the measurement procedure for the signals is planned to be carried out simultaneously with the performance of various exercises, while painful stimuli are administered. Before conducting the measurements with the devices, an analysis of the type of data to be collected will be performed using two open databases containing EEG and fNIRS signals recorded during various painful stimuli. At the time of abstract submission, our work is still in progress. We've conducted an analysis within the realm of signals from open databases, leading to preliminary findings. These initial results have demonstrated significant promise in terms of the ease and precision of signal recording. Notably, we've identified the presence of pain across the alpha, beta, and theta waves, encompassing frequencies spanning from 8 to 70 Hz. Upon applying a painful stimulus, signals within the identified frequencies show a 2 fold increase in amplitude. The expected outcomes are 1)A framework to perform real-time analysis of EEG and fNIRS signals. 2)A neurofeedback system for real-time pain detection. In conclusion, portable EEG and fNIRS devices show promise in detecting and characterizing moments of pain. The analysis presented in the preliminary results, along with the measurement devices used, holds significant potential for conducting real-time analysis in future stages of the research. Furthermore, there is the potential to apply various artificial intelligence models during different exercises. [1]V. Fusetti et al., “Clown therapy for procedural pain in children: a systematic review and meta-analysis,” [2]G. Forte, G. Troisi, M. Pazzaglia, V. De Pascalis, and M. Casagrande, “Heart Rate Variability and Pain: A Systematic Review,” . [3]X. yi Wang et al., “Evaluation of the effect of physical therapy on pain and dysfunction of knee osteoarthritis based on fNIRS: a randomized controlled trial protocol,” [4]N. Birch, J. Graham, C. Ozolins, K. Kumarasinghe, and F. Almesfer, “Home-Based EEG Neurofeedback Intervention for the Management of Chronic Pain,”