Video Experimental Relacionado
Updated: Mar 2, 2026
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
Más allá del bombo: Adopción y actitudes hacia la IA generativa entre los físicos médicos indonesios
L E Lubis1, H D R Raharja2, G M Fajri3
1Department of Physics, Faculty of Mathematics and Natural Sciences, Universitas Indonesia, Kampus UI Depok, Depok 16424, Indonesia; Radiology Unit, Universitas Indonesia Hospital, Kampus UI Depok, Depok 16424, Indonesia.
Purpose:
To evaluate adoption patterns and attitudes toward generative AI-particularly large language models (LLMs) such as ChatGPT-among medical physicists in Indonesia.
Methods:
A cross-sectional online survey was conducted in October 2025 among 165 Indonesian medical physicists. The questionnaire captured demographics, awareness and use of LLMs, professional use cases, perceived productivity benefits, verification behaviors, concerns, and views on future integration. Descriptive statistics were used for analysis.
Results:
LLM awareness was nearly universal (95%), and 96% had used an LLM. Most respondents (72%) used LLMs at least several times per week. Common applications included literature search and summarization (67%), solving technical or clinical questions (58%), drafting professional communications (46%), preparing teaching materials (41%), and assisting scientific writing (39%). Perceived benefits were strong: over 85% reported increased perceived efficiency, with 61% noting significant productivity gains and none reporting decreased productivity. Trust remained moderate; only 11% expressed high confidence in LLM accuracy, and the majority routinely verified important outputs. Key concerns included potential inaccuracies or hallucinations (63%), overreliance (59%), limited domain understanding (59%), ethical or plagiarism risks (55%), and data privacy (44%). Despite these concerns, 94% expected a positive future impact of LLMs on the profession, 79% supported incorporating AI/LLM competencies into training, and 92% would recommend LLM use to colleagues, albeit with caution.
Conclusions:
Indonesian medical physicists have rapidly integrated generative AI into research, education, and problem-solving activities, reporting perceived efficiency gains while maintaining critical oversight. Their optimism underscores the need for guidance and training to ensure responsible, effective use of AI tools in medical physics. Cautions should be made on ethical issues of an overuse/abuse of generative AI especially among early-career or in-training medical physicists.
