Biologie de la Reproduction, Hospices Civil de Lyon, HFME, Bron, France.
Inserm U1208, Bron, France.
Syst Biol Reprod Med. 2021 Feb;67(1):64-78. doi: 10.1080/19396368.2020.1822953.
The aim of this work was o construct a score issued from a machine learning system with self-improvement capacity able to predict the fate of an ART embryo incubated in a time lapse monitoring (TLM) system. A retrospective study was performed. For the training data group, 110 couples were included and, 891 embryos were cultured. For the global setting data group, 201 couples were included, and 1186 embryos were cultured. No image analysis was used; morphokinetic parameters from the first three days of embryo culture were used to perform a logistic regression between the cell number and time. A score named DynScore was constructed, the prediction power of the DynScore on blastocyst formation and the baby delivery were tested via the area under the curve (AUC) obtained from the receiver operating characteristic (ROC). In the training data group, the DynScore allowed the blastocyst formation prediction (AUC = 0.634, p < 0.001), this approach was the higher among the set of the tested scores. Similar results were found with the global setting data group (AUC = 0.638, p < 0.001) and it was possible to increase the AUC of the DynScore with a regular update of the prediction system by reinforcement, with an AUC able to reach a value above 0.9. As only the best blastocysts were transferred, none of the tested scores was able to predict delivery. In conclusion, the DynScore seems to be able to predict the fate of an embryo. The reinforcement of the prediction system allows maintaining the predictive capacity of DynScore irrespective of the various events that may occur during the ART process. The DynScore could be implemented in any TLM system and adapted by itself to the data of any ART center. ART: assisted reproduction technology; TLM: time lapse monitoring system; AUC: area under the curve; ROC: receiver operating characteristic; eSET: elective single embryo transfer; AIS: artificial intelligence system; KID: known implantation data; AMH: anti-Müllerian hormone; BMI: body mass index; WHO: World Health Organization; c-IVF: conventional in-vitro fertilization; ICSI: intracytoplasmic sperm injection; PNf: pronuclear formation; D3: day 3; D5: day 5; D6: day 6; GnRH: gonadotrophin releasing hormone; FSH: follicle stimulating hormone; LH: luteinizing hormone; hCG: human chorionic gonadotropin; PVP: polyvinyl pyrrolidone; PNf: time of pronuclear fading; tx: time of cleavage to x blastomeres embryo; ICM: inner cell mass; TE: trophectoderm; NbCell: number of cells at t time; FIFO: first in first out; TD: training data group; SD: setting data group; R: real world.
这项工作的目的是构建一个具有自我改进能力的机器学习系统评分，该系统能够预测在时间延迟监测（TLM）系统中培养的 ART 胚胎的命运。进行了一项回顾性研究。在训练数据组中，纳入了 110 对夫妇，培养了 891 个胚胎。在全球设置数据组中，纳入了 201 对夫妇，培养了 1186 个胚胎。未使用图像分析；从胚胎培养的头三天使用形态动力学参数在细胞数量和时间之间进行逻辑回归。构建了一个名为 DynScore 的评分，通过接收器操作特性（ROC）获得的曲线下面积（AUC）来测试 DynScore 对囊胚形成和分娩的预测能力。在训练数据组中，DynScore 允许预测囊胚形成（AUC=0.634，p<0.001），这是测试评分中最高的一种。在全球设置数据组中也得到了类似的结果（AUC=0.638，p<0.001），通过强化对预测系统进行定期更新可以提高 DynScore 的 AUC，其 AUC 能够达到 0.9 以上的值。由于仅转移了最好的囊胚，因此没有任何评分能够预测分娩。总之，DynScore 似乎能够预测胚胎的命运。预测系统的强化允许 DynScore 的预测能力保持不变，而不管 ART 过程中可能发生的各种事件。DynScore 可以在任何 TLM 系统中实施，并根据任何 ART 中心的数据进行自我调整。ART：辅助生殖技术；TLM：时间延迟监测系统；AUC：曲线下面积；ROC：接收器操作特性；eSET：选择性单胚胎移植；AIS：人工智能系统；KID：已知植入数据；AMH：抗苗勒管激素；BMI：体重指数；WHO：世界卫生组织；c-IVF：常规体外受精；ICSI：胞浆内精子注射；PNf：原核形成；D3：第 3 天；D5：第 5 天；D6：第 6 天；GnRH：促性腺激素释放激素；FSH：卵泡刺激素；LH：黄体生成素；hCG：人绒毛膜促性腺激素；PVP：聚乙烯吡咯烷酮；PNf：原核消失时间；tx：从受精卵到 x 个卵裂球胚胎的时间；ICM：内细胞团；TE：滋养层；NbCell：t 时间的细胞数；FIFO：先进先出；TD：训练数据组；SD：设置数据组；R：现实世界。
