Jeong Jinseo, Kim Sohyun, Pan Lian, Hwang Daye, Kim Dongseop, Choi Jeongwon, Kwon Yeongkyo, Yi Pyeongro, Jeong Jisoo, Yoo Seok-Ju
College of Medicine, Dongguk University, Gyeongju-si, Republic of Korea.
Department of Preventive Medicine, College of Medicine, Dongguk University, Gyeongju-si, Republic of Korea.
Medicine (Baltimore). 2025 Feb 7;104(6):e41470. doi: 10.1097/MD.0000000000041470.
Artificial intelligence (AI) has revolutionized medical diagnostics by enhancing efficiency, improving accuracy, and reducing variability. By alleviating the workload of medical staff, AI addresses challenges such as increasing diagnostic demands, workforce shortages, and reliance on subjective interpretation. This review examines the role of AI in reducing diagnostic workload and enhancing efficiency across medical fields from January 2019 to February 2024, identifying limitations and areas for improvement. A comprehensive PubMed search using the keywords "artificial intelligence" or "AI," "efficiency" or "workload," and "patient" or "clinical" identified 2587 articles, of which 51 were reviewed. These studies analyzed the impact of AI on radiology, pathology, and other specialties, focusing on efficiency, accuracy, and workload reduction. The final 51 articles were categorized into 4 groups based on diagnostic efficiency, where category A included studies with supporting material provided, category B consisted of those with reduced data volume, category C focused on independent AI diagnosis, and category D included studies that reported data reduction without changes in diagnostic time. In radiology and pathology, which require skilled techniques and large-scale data processing, AI improved accuracy and reduced diagnostic time by approximately 90% or more. Radiology, in particular, showed a high proportion of category C studies, as digitized data and standardized protocols facilitated independent AI diagnoses. AI has significant potential to optimize workload management, improve diagnostic efficiency, and enhance accuracy. However, challenges remain in standardizing applications and addressing ethical concerns. Integrating AI into healthcare workforce planning is essential for fostering collaboration between technology and clinicians, ultimately improving patient care.
人工智能（AI）通过提高效率、提升准确性和减少变异性，彻底改变了医学诊断。通过减轻医务人员的工作量，人工智能应对了诸如诊断需求增加、劳动力短缺以及依赖主观解释等挑战。本综述考察了2019年1月至2024年2月期间人工智能在减少诊断工作量和提高各医学领域效率方面的作用，确定了局限性和改进领域。使用关键词“人工智能”或“AI”、“效率”或“工作量”以及“患者”或“临床”在PubMed上进行的全面搜索共识别出2587篇文章，其中51篇被纳入综述。这些研究分析了人工智能对放射学、病理学和其他专业的影响，重点关注效率、准确性和工作量减少。最终的51篇文章根据诊断效率分为4组，其中A类包括提供了支持材料的研究，B类由数据量减少的研究组成，C类专注于独立人工智能诊断，D类包括报告了数据减少但诊断时间无变化的研究。在需要熟练技术和大规模数据处理的放射学和病理学领域，人工智能提高了准确性，并将诊断时间减少了约90%或更多。特别是放射学领域，C类研究的比例很高，因为数字化数据和标准化协议便于独立人工智能诊断。人工智能在优化工作量管理、提高诊断效率和提升准确性方面具有巨大潜力。然而，在标准化应用和解决伦理问题方面仍存在挑战。将人工智能纳入医疗劳动力规划对于促进技术与临床医生之间的合作至关重要，最终可改善患者护理。
