Related Experiment Video
Updated: Jun 13, 2026
10:14
Author Spotlight: Enhancing Neurorehabilitation Through EEG, Motor Imagery, and Virtual Reality
Published on: May 10, 2024
1.7K
Artificial Intelligence Physician Avatars for Patient Education: A Pilot Study.
Syed Ali Haider1, Srinivasagam Prabha1, Cesar Abraham Gomez-Cabello1
1Division of Plastic Surgery, Mayo Clinic, 4500 San Pablo Rd, Jacksonville, FL 32224, USA.
Journal of Clinical Medicine
|December 11, 2025
Summary
AI-generated surgeon avatars were highly accepted by patients for postoperative education, demonstrating excellent usability and trust. Transparency about the synthetic nature of the technology enhanced patient confidence in these digital physician tools.
Area of Science:
- Artificial Intelligence in Healthcare
- Medical Education Technology
- Human-Computer Interaction
Background:
- Generative AI and synthetic media enable realistic human Embodied Conversational Agents (ECAs), including "digital twins" of physicians.
- These digital twins offer potential for scalable, 24/7 patient communication and education outside clinical settings.
- This pilot study investigated surgical patient perceptions of an AI-generated surgeon avatar for postoperative education.
Purpose of the Study:
- To evaluate surgical patient perceptions of an AI-generated surgeon avatar.
- To assess usability, engagement, trust, eeriness, and realism of the avatar system.
- To gather qualitative feedback on patient experience with AI-driven clinical communication.
Main Methods:
- Pilot feasibility study with 30 plastic surgery patients at Mayo Clinic.
- Developed an interactive surgeon avatar using the HeyGen IV model for likeness replication.
- Patients interacted via voice queries mapped to pre-recorded responses; perceptions assessed using validated scales and qualitative feedback.
Main Results:
- The avatar system achieved 99% query response accuracy, with excellent usability (SUS score 87.7) and high engagement (4.27).
- 100% of participants found the avatar trustworthy and its information believable; eeriness was minimal (1.57).
- 86.6% recognized the avatar as their surgeon, though 70% noted voice resemblance limitations; prior deepfake exposure correlated with higher acceptance.
Conclusions:
- AI-generated physician avatars achieved high patient acceptance without inducing the uncanny valley effect.
- Transparency regarding the avatar's synthetic nature positively impacted trust.
- Synthetic physician avatars, when implemented transparently, represent a scalable solution for postoperative education, potentially preserving trust in physician-patient relationships.
