Lin-Martore Margaret, Kornblith Aaron, Firnberg Maytal, Haque Adnan, O'Brien Bridget
Departments of Emergency Medicine and Pediatrics, University of California, San Francisco, California, USA.
Department of Medicine, University of California, San Francisco, California, USA.
J Am Coll Emerg Physicians Open. 2025 Jun 5;6(4):100173. doi: 10.1016/j.acepjo.2025.100173. eCollection 2025 Aug.
Artificial intelligence (AI) may improve many aspects of point-of-care ultrasound (POCUS) for physicians. However, adoption of AI relies on physician trust. Our study seeks to understand factors influencing physician trust in AI-augmented POCUS.
From November 2023 to April 2024, we conducted semistructured interviews with academic pediatric emergency medicine (PEM) physicians who use POCUS. The interview guide was sensitized by Yang's proposed framework for user trust in AI and explored participants' perspectives on AI-augmented POCUS. We used template analysis to identify themes. Interviews continued until thematic sufficiency was achieved.
We interviewed 14 PEM physicians across career stages with varying POCUS experience. Participants named several specific aspects of POCUS where AI would be beneficial including image acquisition, image interpretation, and workflow enhancement. We identified themes related to 5 factors perceived as influential in physicians' trust in AI-augmented POCUS: (1) technological-AI-augmented POCUS must be reliable, accurate, transparent, and overridable by the physician; (2) contextual-The technology should be generalizable to the clinical population; (3) user-Each physician's clinical experience, comfort with POCUS, and experiences with technology affect their trust; (4) social/organizational-Endorsements by known colleagues, institutions, and national groups can engender trust although the organization or company creating the AI may have variable influence; (5) environmental-AI is everywhere, and there is subconscious use and acceptance.
Technological, contextual, user-related, social, organizational, and environmental factors influenced emergency physician trust in AI-augmented POCUS. Understanding these factors is important for developing AI-augmented POCUS tools.
人工智能（AI）可能会改善医生在床旁超声检查（POCUS）中的多个方面。然而，AI的采用依赖于医生的信任。我们的研究旨在了解影响医生对AI增强型POCUS信任的因素。
2023年11月至2024年4月，我们对使用POCUS的学术儿科急诊医学（PEM）医生进行了半结构化访谈。访谈指南受杨提出的用户对AI信任的框架启发，并探讨了参与者对AI增强型POCUS的看法。我们使用模板分析来确定主题。访谈持续进行，直至达到主题饱和。
我们采访了14名处于不同职业阶段、具有不同POCUS经验的PEM医生。参与者指出了POCUS中AI可能有益的几个具体方面，包括图像采集、图像解读和工作流程优化。我们确定了与被认为对医生对AI增强型POCUS的信任有影响的5个因素相关的主题：（1）技术方面——AI增强型POCUS必须可靠、准确、透明且医生可 override；（2）背景方面——该技术应能推广至临床人群；（3）用户方面——每位医生的临床经验、对POCUS的熟悉程度以及技术体验会影响他们的信任；（4）社会/组织方面——知名同事、机构和国家团体的认可可产生信任，尽管创建AI的组织或公司可能有不同程度的影响；（5）环境方面——AI无处不在，存在潜意识的使用和接受。
技术、背景、用户相关、社会、组织和环境因素影响了急诊医生对AI增强型POCUS的信任。了解这些因素对于开发AI增强型POCUS工具很重要。
