AI-Generated Patient Education: Rapid Development of a Laparoscopic Hysterectomy Recovery Video Using Readily Accessible Tools

Study Objective

To demonstrate the use of Artificial Intelligence (AI) systems to rapidly generate high-quality, accessible and inclusive patient education materials.

Design

Educational video created entirely using inexpensive, web-based AI technologies.

Setting

Large Language Model (LLMs), Diffusion Transformers (DiTs), and Deep Learning Models (DLMs) accessed via web-browser

Patients or Participants

Author only.

Interventions

An interactive, iterative and dynamic process using LLM, DiTs, and DLMs produced a patient education video on laparoscopic hysterectomy recovery. ChatGPT was used to generate a storyboard, scene descriptions and narration script, tailored to video submission guidelines. The narration was voiced using text-to-speech, and visuals were produced using text-to-image and text-to-video platforms.

Measurements and Primary Results

Total AI interaction time was ∼6 hours, with rendering queues ranging from 27 seconds to 4 minutes. Recreating the full video in Spanish, Mandarin, and Hindi took ∼30 minutes, with voice synthesis generating multi-minute blocks in under 15 seconds. Policy restrictions on anatomical images varied unpredictably across platforms. Tools with site-specific editing improved visual accuracy. Prompt engineering improved output quality. Costs were limited to a single paid subscription; all other tools used free access.

Conclusion

AI technologies can empower physicians to bridge financial resource, linguistic, and time barriers to create engaging, accessible and accurate patient education material. With strategic use, these tools offer scalable potential for rapid, multilingual educational material development.

Declaration of AI and AI-Assisted Technologies in the Writing Process Statement: During the preparation of this work the author used ChatGPT, Dall-E, Sora, Kling.ai and Elevenlabs.io in order to enhance accessibility and visual clarity for a non-clinical audience. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.