Time-lapse technology application combined with intellectual intelligence to choose the potential physical experiences in IVF

MSc Le Nhat Quang and Ms. Nguyen Thi Phuong Dung

IVFMD, My Duc Hospital

The evaluation and selection of potential embryos for patient use plays a critical role in the success of in vitro fertilization (IVF). Therefore, these processes are always paid much attention to and improved to increase the efficiency of IVF. Currently, almost all IVF centers in the world evaluate embryos based on morphological characteristics at embryonic development times to select potential embryos. In recent years, embryo culture techniques making use of time-lapse cameras have been increasingly used to better support the development of embryos.

The role and challenge of time-lapse technique in potential embryo selection

The time-lapse technique helps to capture image of embryos during cultivation outside the body. Based on that, an embryologist can precisely analyze the embryo development process in each time period as well as the division of embryonic cells. At the same time, certain negative morphological features such as multicellular embryos, unevenly divided embryos, embryos dividing directly from 1 cell into more than 2 cells or other disorders during cell division are discovered to select better potential embryos.

Exploiting the information of embryos effectively recorded from time-lapse will help to select potential embryos more accurately, improving results at the first embryo transfers. Thus, it is possible to shorten the time to achieve the outcome  - the birth of a baby, improving the cost-effectiveness of a treatment. On that basis, the time-lapse technology can be effective for all IVF patients. . However, with limited resources, this time-lapse technique can be prioritized for older patients (> 35 years old), patients who fail  in embryo implantation multiple times due to unknown causes related to the quality of embryo.

In reality, whether the time-lapse technology can play its intended role or not depends on how effectively we apply and exploit data. Current methods of assessing the embryo quality based on its morphology are highly variable, depending on the personal experience of an embryologist. The time-lapse technique can provide better information for embryologists during the embryo quality assessment. However, embryologists also need a lot of time to analyze and evaluate embryos. Moreover, there are many kinetic parameters as well as morphological characteristics that humans cannot synthesize and analyze at the same time to select the embryos with the highest implantation potential for patients to use. All of these factors create a common challenge of how to effectively exploit the data source about embryos collected by the time-lapse system to select the highest potential embryos.

Clinical effectiveness of time-lapse technique in IVF

Although the time-lapse technique has been routinely applied to IVF since 2010 and has become popular rapidly in recent years, the clinical effectiveness of this technique compared with traditional embryo culture and selection is still waiting for confirmation. A number of randomized controlled trials have shown that the application of time-lapse technology improved the rate of fetal progression and the outcome after IVF compared with the routine embryo culture and selection. The study of Siristatidis et al. (2015) surveyed 239 patients, who were divided into 2 groups: having embryos cultured with the time-lapse technique and having the ones using the conventional culture method. Results were significantly higher regarding clinical pregnancy rates, progression rates and rates of alive babies in the first group  compared to the second group, reaching 65.71% compared to 39.05%; 55.71% compared to 31.36% and 45.71% compared to 28.40%;

How does artificial intelligence combine with time-lapse technology?

Thanks to the strong growth of information technology, artificial intelligence (AI) has been developed and applied in many areas of life, including healthcare. Artificial intelligence is a field in computer science, where  people program machines to simulate human actions. Artificial intelligence is capable of continuous learning from large data sources to gain experience in information processing. As a result, artificial intelligence can systematically process large-scale and complex data in a very fast speed compared to humans.

In order to overcome the weaknesses of assessment and selection of embryos based on morphology or kinematic parameters recorded by time-lapse, artificial intelligence has been successfully researched and applied to handle large quantities of information about embryos and the classification of embryos according to the potential for conception. Accordingly, the image data of embryos from the time-lapse system is aggregated and processed through an artificial intelligence model to classify embryo quality and anticipate the performance of embryo use. At the same time, the algorithms in the artificial intelligence model are automatically set up and continuously updated with the latest data from the system to improve the accuracy of the information processing. Therefore, the artificial intelligence model supports embryologists to select the best embryos to deliver to a patient.

The artificial intelligence technology combined with time-lapse will save time and effort, eliminate subjective factors of embryologists, and improve accuracy in the process of evaluating and selecting embryos. As a result, the task of choosing the most potential embryo for each embryo transfer is improved, limiting the transfer of more than one embryo which causes multiple pregnancy and helping to bring a healthy pregnancy for both mother and baby.  This improvement in the success rate of IVF helps reduce time and cost for patients.

Some works applying artificial intelligence in selecting potential embryos

In 2018, Dr. Tran Dang Dinh Ang et al (University of New South Wales - Australia) published a study where IVY was used to predict the pregnancy rate of an embryo via data analysis of a time-lapse system . Accordingly, IVY is an artificial intelligence model built from the analysis of video records of an embryonic development through the time-lapse technology and the comparison with the fetal outcome of that embryo. Research results show that IVY is capable of accurately predicting the conception potential of an embryo in 89% of the cases.

Another project launched by a group of Australian Doctors including Jonathan Hall, Michelle Perugini and Don Perugini in 2016 with the application "Life Whisperer". This application analyzes 2D embryonic image data to predict pregnancy. The results announced at ESHRE 2019 show that Life Whisperer has the ability to select embryos 30.8% more accurately than embryologists do based on morphological criteria.

Currently, many other research projects on the application of artificial intelligence in selecting embryos have been launched around the world. In Vietnam, as the number of IVF cycles is rising and the time-lapse technology is increasingly applied, more and more embryonic image data can be collected to become a basis for conducting research on the applications of artificial intelligence in potential embryo selection. The time-lapse technology combined with artificial intelligence is and will be able to bring about positive changes in the field of IVF.

References

1. Armstrong S, Bhide P, Jordan V, Pacey A, Marjoribanks J, Farquhar C. 2019. Time-lapse systems for embryo incubation and assessment in assisted reproduction. Cochrane Database of Systematic Reviews. Issue 5. Art. No.: CD011320.

2. Pribenszky C, Nilselid AM, Montag M. 2017. Time-lapse culture with morphokinetic embryo selection improves pregnancy and live birth chances and reduces early pregnancy loss: a meta-analysis. Reprod Biomed Online.35(5):511-520.

3. Quang, L. N., Tram, L. T. B., Thuy, N. H. M., Toan, P. D., Vinh, D. Q., & Huyen, N. T. T. 2018. Comparison of clinical outcome of frozen embryo transfer after embryo selection based on morphokinetic versus morphologic criteria for freezing. Biomedical Research and Therapy, 5(12), 2910-2917.

4. Siristatidis, C., Komitopoulou, M. A., Makris, A., Sialakouma, A., Botzaki, M., Mastorakos, G., Salamalekis, G., Bettocchi, S. & Palmer, G. A. 2015. Morphokinetic parameters of early embryo development via time lapse monitoring and their effect on embryo selection and ICSI outcomes: a prospective cohort study. J Assist Reprod Genet, 32, 563-70.

5. Tran, A., Cooke, S., Illingworth, P. J., & Gardner, D. K. (2018). Artificial intelligence as a novel approach for embryo selection. Fertility and Sterility, 110(4), e430.

6. Wu, L., Han, W., Wang, J., Zhang, X., Liu, W., Xiong, S., Han, S., Liu, J., Gao, Y. & Huang, G. 2017. Embryo culture using a time-lapse monitoring system improves live birth rates compared with a conventional culture system: a prospective cohort study. Hum Fertil (Camb), 1-8.