Automatic detection of morphodynamic profile of human embryos in vitro data sets preparation
Author(s) -
G.B. Nemkovskiy,
О. В. Шурыгина,
A.A. Bayzarova,
Dmitry Rusakov,
A. B. Kuznetsov,
V.K. Belyakov
Publication year - 2020
Publication title -
procedia computer science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.334
H-Index - 76
ISSN - 1877-0509
DOI - 10.1016/j.procs.2020.09.212
Subject(s) - embryo , embryo transfer , blastocyst , computer science , biology , andrology , embryogenesis , medicine , genetics
The article defines "Human embryo‘s morphodynamic profile" and a set of markers of its components. The cultivation of human embryos in vitro in the practice of embryological laboratories is currently a well-established and standardized technique. Nevertheless, the problem of determining reliable predictors of a developing embryo, which has the highest chances of implantation, is extremely urgent. Currently, the concepts of "morphokinetics" or "morphokinetic state" determine the apparent fixed state of a human embryo. The sequential change in morphokinetic states over time is the essence of morphodynamics. To standardize the description of the development of in vitro cultured human embryos, we introduced the concept of "Morphodynamic profile of a human embryo." This definition will allow specialists to more reasonably carry out the ranking of cultured embryos for the purpose of selection for transfer into the uterine cavity and cryopreservation. The goal is the determination of key morphokinetic parameters (predictors) of embryo’s cultivation process. The study was performed in the laboratory of ART (assisted reproductive technologies) of the Clinical Hospital IDK (group of companies "Mother and Child"). For non-invasive monitoring of pre-implantation cultivation of human embryos up to the blastocyst stage (0-6 days from insemination), a multigas incubator with time-lapse EmbryoVisor system (WESTTRADE LTD, Russia) was used. To test the technique, the morphodynamic profiles of 672 human embryos were formed considering the arrangement of markers of fixed morphokinetic states. Graphical data sets were selected and downloaded for convolutional neural network training. The data obtained, supplemented by the results of transfers and pregnancy outcomes, will allow us to develop a subsystem for the automated recognition of the morphokinetic states of a human embryo and the automatic assessment of their implantation ability.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom