Chen Yimin, Hu Xi, Liu Yang
（ 650101） Department of Reproductive Medicine, The Second Affiliated Hospital of Kunming Medical University, Kunming 650101, China.
Sichuan Da Xue Xue Bao Yi Xue Ban. 2025 Mar 20;56(2):563-570. doi: 10.12182/20250360401.
In recent years, with the constant rise in the number of infertility patients, the rapid and accurate assessment of sperm quality has become a key challenge in reproductive medicine. Currently, traditional methods for sperm classification suffer from limitations in efficiency, subjectivity, and cost-effectiveness, and hence fail to meet clinical needs. With the introduction of artificial intelligence (AI), innovative approaches have been offered to address these issues. Herein, we systematically reviewed the latest progress in AI applications in the field of male infertility, focusing on the role of AI in the assessment of sperm concentration, motility, morphology, DNA fragmentation index, and the treatment of non-obstructive azoospermia. It has been reported that AI technologies, such as convolutional neural networks, demonstrate high accuracy and efficiency in assessing sperm concentration and motility. Particularly in morphological analysis, the performance of AI has been validated in multiple studies, significantly enhancing objectivity and clinical utility. However, the assessment of DNA fragmentation index remains underexplored due to the lack of support in advanced imaging technology, with only a few models showing promise. Additionally, AI significantly improves sperm detection rates in modified testicular sperm extraction through AI-driven image recognition, offering a breakthrough in the treatment of patients with non-obstructive azoospermia. We also discussed the application of natural language processing technology in patient pre-consultation and follow-up, such as automated data collection and intelligent tracking systems, which demonstrate AI's potential to optimize medical workflows. In the future, with the accumulation of high-quality datasets, algorithm optimization, and advances in imaging technology, the application of artificial intelligence is expected to enable multi-dimensional comprehensive screening and play a greater role in the diagnosis and treatment of male infertility.
近年来，随着不孕患者数量的不断增加，精子质量的快速准确评估已成为生殖医学中的一项关键挑战。目前，传统的精子分类方法在效率、主观性和成本效益方面存在局限性，因此无法满足临床需求。随着人工智能（AI）的引入，人们提出了创新方法来解决这些问题。在此，我们系统回顾了人工智能在男性不育领域应用的最新进展，重点关注人工智能在精子浓度、活力、形态、DNA碎片指数评估以及非梗阻性无精子症治疗中的作用。据报道，卷积神经网络等人工智能技术在评估精子浓度和活力方面具有很高的准确性和效率。特别是在形态学分析方面，人工智能的性能在多项研究中得到了验证，显著提高了客观性和临床实用性。然而，由于缺乏先进成像技术的支持，DNA碎片指数的评估仍未得到充分探索，只有少数模型显示出前景。此外，通过人工智能驱动的图像识别，人工智能显著提高了改良睾丸精子提取中的精子检测率，为非梗阻性无精子症患者的治疗带来了突破。我们还讨论了自然语言处理技术在患者预咨询和随访中的应用，如自动数据收集和智能跟踪系统，这些展示了人工智能优化医疗工作流程的潜力。未来，随着高质量数据集的积累、算法优化和成像技术的进步，人工智能的应用有望实现多维度综合筛查，并在男性不育的诊断和治疗中发挥更大作用。
