Educational impact and cost efficiency of AI-enhanced videos in pediatric surgery training: a quasi-experimental study

Abstract

Pediatric surgery training is challenged by declining clinical exposure and limited access to high-quality educational resources. Advances in generative artificial intelligence (AI) allow enhancement of educational videos without professional studio infrastructure, but evidence supporting their educational value remains limited. To evaluate the educational impact and cost efficiency of locally deployed AI-enhanced lecture videos compared with standard faculty-recorded videos in pediatric surgery training. A quasi-experimental, non-randomized, post-test-only study conducted at Al-Azhar University, Cairo, Egypt between December 2024 and September 2025. Medical students and pediatric surgery residents viewed standardized lecture videos on five core pediatric surgery topics. Identical scripts and narration were used in both groups; the intervention differed only by AI-based background enhancement using open-source, locally deployed tools. Primary outcomes were perceived educational clarity (five-point Likert scale) and knowledge acquisition assessed by a 40-item multiple-choice examination. Secondary outcomes included visual comfort, perceived professionalism, and willingness to recommend the format. Between-group comparisons and effect sizes were calculated. A cost analysis compared AI-enhanced production with traditional studio-based methods. A total of 240 participants were included (120 per group). AI-enhanced videos were associated with higher educational clarity scores (mean 4.52 vs. 3.88; p < 0.001) and higher knowledge scores (76.8% vs. 69.5%; p < 0.001). Secondary outcomes consistently favored AI-enhanced videos, including visual comfort and perceived professionalism. Locally deployed AI production reduced per-video costs by 72–94% compared with traditional studio production. Locally deployed AI-enhanced videos improved short-term educational outcomes and substantially reduced production costs while preserving presenter authenticity. This approach represents a practical, scalable strategy for enhancing pediatric surgery education, particularly in resource-limited settings. Further studies are needed to assess long-term retention and clinical impact.

Ähnliche Arbeiten