RE-HPBS-IPIC: A Resting EEG- and High-Activation Pain Brain Source-Driven Framework for Inter-Subject Pain Intensity...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Machine Learning

Background:

• Accurate inter-subject pain intensity assessment using electroencephalography (EEG) is challenging due to significant individual differences in brain activity.
• Existing methods struggle to reliably classify pain intensity across different individuals.

Purpose of the Study:

• To develop a novel framework for reliable inter-subject pain intensity classification using EEG.
• To leverage pain-related brain dynamics and transfer learning to overcome inter-subject variability.

Main Methods:

• Quantified pain sensitivity from resting-state EEG to identify subjects with similar neural pain signatures.
• Localized and remapped high-activation pain brain sources between subjects.
• Employed balanced distribution adaptation to align brain source features and mitigate variability.
• Used transfer learning to infer pseudo-labels for target EEG data and guide pain response extraction.

Main Results:

• The proposed framework significantly outperformed three existing approaches in inter-subject pain intensity classification.
• Experimental evaluations on real EEG datasets validated the method's effectiveness.

Conclusions:

• The developed method effectively addresses the challenge of poor reliability in inter-subject pain intensity classification.
• This provides a robust and clinically viable solution for objective pain assessment.