代谢成像用于配子和胚胎评估:人工智能的新途径?
本文综述无标记代谢成像通过NAD(P)H和FAD自发荧光评估配子与胚胎发育潜能,涉及荧光寿命、高光谱和光片显微镜等技术,并探讨其与人工智能结合用于胚胎选择决策支持的前景。
要点
- NAD(P)H/FAD自发荧光可光学反映氧化还原平衡与线粒体活性。
- 代谢特征与生殖衰老、发育潜能及胚胎倍性状态相关。
- 代谢成像结合AI可能实现自动化、客观的配子与胚胎选择。
核心结论
- 代谢成像尤其结合人工智能,可能为配子和胚胎选择提供自动化、客观的决策支持。
研究设计与样本
- 受邀小型综述,综合NAD(P)H/FAD自发荧光生化基础及生殖相关成像研究。
主要发现
- NAD(P)H/FAD自发荧光提供氧化还原平衡、线粒体活性和氧化代谢的光学替代指标。
- 代谢特征与生殖衰老、发育潜能及胚胎倍性状态相关。
临床怎么用
- 代谢成像可作为形态学和遗传学方法的补充,用于评估配子与胚胎发育潜能。
- 结合AI可将复杂代谢成像数据转化为胚胎排序和风险分层的临床决策支持输出。
局限性
- 不同成像模态存在信号偏倚,标准化仍需共识推进。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
Infertility affects approximately one in six people globally and demand for ARTs, is expected to rise as parenthood is increasingly delayed. However, ART success remains constrained by gamete and embryo quality, female age, and biological factors beyond chromosomal status alone, highlighting the need for non-invasive methods that can complement current morphology-based and genetic approaches. Because gamete and embryo developmental competence is tightly coupled to cellular metabolism, label-free metabolic imaging has emerged as a promising strategy to assess developmental potential through endogenous autofluorescence of reduced nicotinamide adenine dinucleotide/phosphate [NAD(P)H] and oxidized flavins (FAD), which provide optical proxies of redox balance, mitochondrial activity, and oxidative metabolism. This invited mini-review synthesizes the biochemical basis of NAD(P)H/FAD autofluorescence signals, relates these readouts to the unique metabolic programs of oocytes, embryos across preimplantation development, and sperm, and reviews reproductive studies using fluorescence lifetime imaging microscopy, hyperspectral microscopy, and emerging light-sheet fluorescence microscopy. We discuss practical and interpretive challenges, including modality-dependent signal biases and recent consensus efforts towards standardization. Evidence linking metabolic signatures to reproductive ageing, developmental potential, and embryo ploidy status suggests that metabolic imaging, particularly when paired with artificial intelligence (AI), could enable automated, objective decision support for gamete and embryo selection. Finally, we briefly outline how AI could convert complex metabolic imaging data into clinically interpretable decision-support outputs for embryo ranking and risk stratification. Advances in rapid volumetric imaging, microsystems, and AI may support future ART workflows aimed at improving efficiency, accessibility, and cost-effectiveness.
引用
- Horta F, Sakkas D, Handley S, et al. Metabolic imaging for gamete and embryo assessment through advanced microscopy technologies: a novel avenue for artificial intelligence? Hum Reprod. 2026. PMID: 42449562. doi:10.1093/humrep/deag114
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。