Assessing ChatGPT-4o's Image Generation for Educational Infographics in Plastic Surgery: A Pilot Study.

[BACKGROUND] Integrating generative artificial intelligence (AI) in patient education has opened new avenues for creating engaging and informative content, especially in visually intensive specialties such as plastic surgery. ChatGPT-4o, developed by OpenAI, incorporates image generation capabilities alongside advanced language processing, offering the potential for rapid development of illustrated educational tools tailored to patient needs.

[METHODS] This pilot study explored the efficacy of ChatGPT-4o in generating educational infographics for five commonly performed plastic surgery procedures: rhinoplasty, blepharoplasty, abdominoplasty, facelift, and liposuction. For each procedure, structured prompts were used to create infographics that combined explanatory text with detailed visuals. The outputs were evaluated by a panel of expert plastic surgeons (N = 4) across five domains: visual appeal, clarity of information, anatomical accuracy, educational value, and overall suitability for patient education. Evaluation scores were analyzed using descriptive statistics and one-sample t tests.

[RESULTS] All infographics received mean Likert scores exceeding 4.0 in each domain, with visual appeal rated highest (mean 4.72 ± 0.08) and anatomical accuracy comparatively lower (mean 4.32 ± 0.08). Statistical analysis confirmed that the differences from the benchmark of 4.0 were significant in all domains (p < 0.001), supporting the model's effectiveness. As no real patient data or identifiable information was used, the study was conducted in accordance with the Declaration of Helsinki, and no institutional ethics review was required.

[CONCLUSION] ChatGPT-4o demonstrates strong potential as a tool for generating educationally effective and visually appealing infographics for patient education in plastic surgery. While further refinement is needed to improve anatomical precision, these findings support broader integration of AI-generated visual tools into clinical practice, with implications for improving patient communication, engagement, and informed consent.

[LEVEL OF EVIDENCE IV] This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .