Sciorio Romualdo, Tramontano Luca, Gullo Giuseppe
Fertility Medicine and Gynaecological Endocrinology Unit, Department Woman Mother Child, Lausanne University Hospital, 1011 Lausanne, Switzerland.
Département de Gynécologie-Obstétrique, Réseau Hospitalier Neuchâtelois, 2000 Neuchâtel, Switzerland.
JBRA Assist Reprod. 2025 Jul 30;29(2):338-350. doi: 10.5935/1518-0557.20250019.
During human in-vitro culture, morphological microscope analysis is routinely used to select embryos with the highest implantation potential for transfer, aiming for successful pregnancy and healthy live birth. This evaluation includes blastomere number, size, fragmentation, multinucleation, blastocyst (BL) expansion, and the inner-cell mass and trophectoderm appearance. However, this method requires removing embryos from the incubator, exposing them to non-physiological conditions such as fluctuations in pH, temperature, gases concentrations, as well as significant inter-observer variability. Continuous embryo culture using time-lapse monitoring (TLM) has revolutionized embryo evaluation by allowing continuous, real-time tracking of embryo development from fertilisation to blastocyst formation. This reduces the need to remove embryos from the incubator and helps maintain stable culture conditions. The monitoring system typically includes a standard incubator with an integrated microscope coupled to a digital camera, capturing images at regular intervals that are processed into a video for analysis. Despite its advantages, accurately predicting implantation rates in humans remains challenging. Recently, artificial intelligence (AI) has emerged as promising tool to objectively evaluate human embryos. AI can analyse large datasets, including embryological, clinical, and genetic information, and assist in individualizing treatment protocols. Integrating AI with TLM could improve embryo selection and enhance overall success rates. This paper explores the potential benefits of combining TLM and AI in reproductive and embryology laboratories, highlighting their potential to improve the outcomes of human ART.
在人类体外培养过程中,形态学显微镜分析通常用于选择具有最高着床潜力的胚胎进行移植,目标是实现成功妊娠和健康活产。这种评估包括卵裂球数量、大小、碎片率、多核化、囊胚扩张以及内细胞团和滋养外胚层的外观。然而,这种方法需要将胚胎从培养箱中取出,使其暴露于非生理条件下,如pH值、温度、气体浓度的波动,以及观察者之间存在显著差异。使用延时监测(TLM)进行连续胚胎培养彻底改变了胚胎评估方式,它允许从受精到囊胚形成对胚胎发育进行连续、实时跟踪。这减少了将胚胎从培养箱中取出的需求,并有助于维持稳定的培养条件。监测系统通常包括一个标准培养箱,该培养箱配有一台与数码相机相连的显微镜,定期拍摄图像并处理成视频进行分析。尽管有其优点,但准确预测人类的着床率仍然具有挑战性。最近,人工智能(AI)已成为客观评估人类胚胎的有前途的工具。人工智能可以分析大型数据集,包括胚胎学、临床和遗传信息,并协助制定个性化的治疗方案。将人工智能与延时监测相结合可以改善胚胎选择并提高总体成功率。本文探讨了在生殖和胚胎学实验室中将延时监测和人工智能相结合的潜在益处,强调了它们改善人类辅助生殖技术结果的潜力。