Decoding of Pain Perception using EEG Signals for a Real-Time Reflex System in Prostheses: A Case Study

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Medicine

Background:

• Restoring sensory perception, including touch and pain, is crucial for upper-limb amputees to regain embodiment and protection.
• While tactile restoration has advanced, understanding and decoding pain perception dynamics remain significant challenges.

Purpose of the Study:

• To identify and validate spatio-temporal electroencephalography (EEG) signatures of brain activity during different levels of phantom limb stimulation in an upper-limb amputee.
• To develop and test a system for detecting pain perception and reaction based on EEG features.
• To investigate cortical activity patterns associated with innocuous, moderately intense, and noxious stimuli.

Main Methods:

• Utilized electroencephalography (EEG) recordings during transcutaneous nerve stimulation (TENS) of a phantom limb in an upper-limb amputee.
• Developed a classification system based on spatio-temporal EEG features to differentiate between stimulation conditions.
• Analyzed cortical activity and localized sources in response to varying sensory stimuli.

Main Results:

• A classification system accurately distinguished three stimulation conditions (innocuous, moderate, noxious) with 94.66% test accuracy.
• Noxious stimulation showed peak activation in the pre-motor and central cortex (Cz) between 450-750 ms post-stimulation.
• Moderately intense stimulation activated the parietal lobe, while noxious stimuli triggered early activation of the anterior cingulate cortex (ACC) and posterior cingulate cortex (PCC).

Conclusions:

• This study presents a novel approach to analyzing and classifying neural activity for sensory restoration in amputees.
• The findings demonstrate the potential for real-time pain detection systems in upper-limb prostheses.
• Although a single case study, it offers a foundational method for future research in neuroprosthetics and pain perception.