Related Experiment Video
Updated: Jun 28, 2026

12:33
Corticospinal Excitability Modulation During Action Observation
Published on: December 31, 2013
9.0K
Exploring EEG Responses during Observation of Actions Performed by Human Actor and Humanoid Robot
Anh T Nguyen1, Ajay Anand2, Michelle J Johnson3
1Anh T. Nguyen is with the School of Science and Applied Science, Department of Bioengineering, University of Pennsylvania, Philadelphia, PA, USA.
Summary
Humanoid robots show potential in supporting action observation (AO) therapy for neurological rehabilitation. Brain activity analysis using electroencephalography (EEG) revealed common neural network activation during AO, regardless of the actor.
Area of Science:
- Neuroscience
- Rehabilitation Robotics
- Human-Robot Interaction
Background:
- Action Observation (AO) therapy is a key rehabilitative approach for motor and language recovery post-neurological conditions like stroke.
- The integration of robotics into rehabilitation settings offers novel avenues for therapeutic interventions.
- Understanding neural responses to observed actions is crucial for optimizing AO therapy protocols.
Purpose of the Study:
- To pilot the use of humanoid robots in supporting AO therapy.
- To investigate neural activity differences when observing actions performed by humans versus robots.
- To assess the feasibility of using electroencephalography (EEG) for monitoring brain responses during robot-assisted AO.
Main Methods:
- Three healthy participants underwent electroencephalography (EEG) monitoring.
- Participants observed actions performed by a human actor and a humanoid robot using their left and right arms.
- Event-Related Spectral Perturbations (ERSPs) in sensorimotor regions were analyzed to assess brain activity.
Main Results:
- Individual differences in ERSP patterns were observed, including suppression in sensorimotor mu and beta rhythms.
- One participant exhibited heightened neural responses to robot-observed actions compared to human-observed actions.
- Strong correlations in ERSPs across all conditions suggest shared neural network engagement within the mirror neuron system during AO.
Conclusions:
- The findings support the feasibility of employing EEG to differentiate neural responses to human and robot actions in AO therapy.
- Humanoid robots show promise as tools to augment AO therapy in rehabilitation.
- Further research is warranted to explore the therapeutic potential of robot-guided AO.

