Herbawani Lindya Okti, Susanti Ari Indra, Adnani Qorinah Estiningtyas Sakilah
Master of Midwifery Study Program, Padjadjaran University, Bandung, West Java, Indonesia.
Department of Public Health, Faculty of Medicine, Padjadjaran University, Bandung, West Java, Indonesia.
Adv Med Educ Pract. 2025 Aug 30;16:1579-1599. doi: 10.2147/AMEP.S543098. eCollection 2025.
Currently, midwifery education is confronted with a variety of obstacles, such as inadequate resources and conventional learning methods that are less effective in enhancing the clinical skills of students. Technological advancements and the rapid evolution of maternal and neonatal health services necessitate the transformation of midwifery education to a competency-based curriculum and outcome-based assessment paradigm. Artificial intelligence (AI) and deep learning have the potential to provide adaptive, personalized, and precise learning in this context. Nevertheless, its implementation continues to encounter a variety of challenges.
This study reviews the role of AI and deep learning algorithms in enhancing outcome-based assessments in midwifery education, focusing on improvements in objectivity, personalized learning, and students' clinical readiness.
This study employed a systematic literature review from Science Direct, Semantic Scholar, Springer Nature, and Taylor and Francis databases. Rayyan's software was employed to select 15 articles from the 771 articles that were discovered, in accordance with the inclusion and exclusion criteria. To guarantee objectivity and quality, two researchers conducted an independent evaluation.
Our review indicates that algorithms including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), Random Forest, and Support Vector Machine (SVM) are proficient in facilitating objective evaluations, delivering tailored feedback, and enhancing clinical learning simulations. Artificial intelligence has demonstrated the capacity to enhance students' communication, critical thinking, and clinical decision-making abilities. The primary challenges encompass infrastructure preparedness, digital literacy, and ethical concerns pertaining to data protection and algorithmic prejudice.
Artificial intelligence and deep learning possess significant promise to revolutionize achievement-based assessments in midwifery education through accurate, adaptable, and scalable evaluations. The successful implementation relies on the management of technological, pedagogical, and ethical restrictions, along with thorough integration into the curriculum to equip graduates for global maternal and neonatal health concerns.
目前，助产士教育面临着各种障碍，如资源不足以及传统学习方法在提高学生临床技能方面效果欠佳。技术进步以及孕产妇和新生儿健康服务的快速发展使得助产士教育需要转变为基于能力的课程和基于结果的评估模式。在这种背景下，人工智能（AI）和深度学习有潜力提供适应性、个性化和精确的学习。然而，其实施仍面临各种挑战。
本研究回顾了人工智能和深度学习算法在加强助产士教育中基于结果的评估方面的作用，重点关注客观性、个性化学习和学生临床准备情况的改善。
本研究对科学Direct、语义学者、施普林格自然和泰勒与弗朗西斯数据库进行了系统的文献综述。根据纳入和排除标准，使用Rayyan软件从发现的771篇文章中筛选出15篇文章。为确保客观性和质量，两名研究人员进行了独立评估。
我们的综述表明，包括卷积神经网络（CNN）、长短期记忆（LSTM）、随机森林和支持向量机（SVM）在内的算法擅长促进客观评估、提供个性化反馈以及增强临床学习模拟。人工智能已证明有能力提高学生的沟通、批判性思维和临床决策能力。主要挑战包括基础设施准备、数字素养以及与数据保护和算法偏见相关的伦理问题。
人工智能和深度学习有很大的潜力通过准确、适应性强和可扩展的评估彻底改变助产士教育中基于成果的评估。成功实施依赖于对技术、教学和伦理限制的管理，以及全面融入课程，以使毕业生能够应对全球孕产妇和新生儿健康问题。
