Temsah Mohamad-Hani, Nazer Rakan, Altamimi Ibraheem, Aldekhyyel Raniah, Jamal Amr, Almansour Mohammad, Aljamaan Fadi, Alhasan Khalid, Temsah Abdulkarim A, Al-Eyadhy Ayman, Aljafen Bandar N, Malki Khalid H
Pediatrics, College of Medicine, King Saud University, Riyadh, SAU.
Pediatric Intensive Care Unit, King Saud University Medical City, Riyadh, SAU.
Cureus. 2025 Jan 17;17(1):e77593. doi: 10.7759/cureus.77593. eCollection 2025 Jan.
The rapid evolution of generative artificial intelligence (AI) has introduced transformative technologies across various domains, with text-to-video (T2V) generation models emerging as transformative innovations in the field. This narrative review explores the potential of T2V AI generation models used in healthcare, focusing on their applications, challenges, and future directions. Advanced T2V platforms, such as Sora Turbo (OpenAI, Inc., San Francisco, California, United States) and Veo 2 (Google LLC, Mountain View, California, United States), both announced in December 2024, offer the capability to generate high-fidelity video contents. Such models could revolutionize healthcare by providing tailored videos for patient education, enhancing medical training, and possibly optimizing telemedicine. We conducted a comprehensive narrative literature search of databases including PubMed and Google Scholar, and identified 41 relevant studies published between 2020 and 2024. Our findings reveal significant possible benefits in improving patient education, standardizing customized medical training, and enhancing remote medical consultations. However, critical challenges persist, including risks of misinformation (or deepfake), privacy breaches, ethical concerns, and limitations in video authenticity. Detection mechanisms for deepfakes and regulatory frameworks remain underdeveloped, necessitating further interdisciplinary research and vigilant policy development. Future advancements in T2V AI generation models could enable real-time healthcare visualizations and augmented reality training. However, achieving these benefits will require addressing accessibility challenges to ensure equitable implementation and prevent potential disparities. By addressing these challenges and fostering collaboration among stakeholders, healthcare systems and AI technologists, T2V AI generation models could transform global healthcare into a more effective, universal, and innovative system while safeguarding against its potential misuse.
生成式人工智能（AI）的快速发展在各个领域引入了变革性技术，文本到视频（T2V）生成模型成为该领域的变革性创新。这篇叙述性综述探讨了医疗保健中使用的T2V AI生成模型的潜力，重点关注其应用、挑战和未来方向。2024年12月宣布的先进T2V平台，如Sora Turbo（美国加利福尼亚州旧金山的OpenAI公司）和Veo 2（美国加利福尼亚州山景城的谷歌有限责任公司），具备生成高保真视频内容的能力。此类模型可为患者教育提供定制视频、加强医学培训并可能优化远程医疗，从而彻底改变医疗保健。我们对包括PubMed和谷歌学术在内的数据库进行了全面的叙述性文献检索，确定了2020年至2024年间发表的41项相关研究。我们的研究结果显示，在改善患者教育、规范定制医学培训和加强远程医疗咨询方面存在显著的潜在益处。然而，关键挑战依然存在，包括错误信息（或深度伪造）风险、隐私泄露、伦理问题以及视频真实性方面的限制。深度伪造检测机制和监管框架仍不发达，需要进一步开展跨学科研究并谨慎制定政策。T2V AI生成模型的未来进展可能实现实时医疗可视化和增强现实培训。然而，要实现这些益处，需要应对可及性挑战，以确保公平实施并防止潜在差距。通过应对这些挑战并促进利益相关者、医疗保健系统和AI技术专家之间的合作，T2V AI生成模型可以将全球医疗保健转变为一个更有效、通用和创新的系统，同时防范其潜在的滥用。
