Time-Resolved Decoding of Perspective and Subjective Pain in EEG

Empathy for pain involves both the initial perception of another person’s suffering and subsequent evaluative processes that shape how that pain is interpreted and experienced. These processes do not operate in isolation but can vary depending on whether individuals imagine themselves in the situation (Self perspective) or focus on the other person’s experience (Other perspective). Traditional EEG-ERP analyses focus on amplitude differences within predefined time windows and electrodes, which may overlook condition-related information distributed across the scalp. The present project applies time-resolved multivariate pattern analysis (MVPA) to examine whether distributed EEG activity encodes stimulus category, instructed perspective, and subjective pain experience.

We will analyze an existing EEG dataset collected using an empathy-for-pain task. Participants viewed painful and non-painful images while adopting either a Self or Other perspective and provided trial-level ratings of painfulness and unpleasantness. EEG was recorded continuously from a 21-channel scalp system during task performance. Analyses will be conducted at the single-trial level using scalp voltage patterns across electrodes at each time point.

We expect distributed EEG patterns to differentiate between painful and non-painful stimuli. We will also test whether neural pattern discriminability varies as a function of instructed perspective. In addition, time-resolved multivariate regression will assess whether distributed EEG activity predicts trial-level subjective ratings. This approach allows to evaluate brain–behavior coupling by determining whether variability in subjective experience is reflected in multivariate neural patterns. Individual differences in trait empathy (IRI) will be examined in relation to neural discriminability and prediction strength. Analyses will be implemented in Python using MNE-Python and scikit-learn.

This project provides a temporally resolved perspective on how empathy-related processes are encoded in distributed neural activity. Using multivariate modeling, we examine how perspective-taking and pain evaluation relate to patterns of neural activity over time.