Page 8 - AIH-2-3

Artificial Intelligence in Health AI in embryo selection for ART

infertility has been decreasing, potentially due to of AI-based ART into the embryo selection process. Key
greater access to infertility treatment facilities. Assisted input data include genetic profiles, historical success
2
reproductive technologies (ARTs) hold promise for rates, and medical histories. Using deep learning and ML
struggling couples with infertility. It can be a common methods, AI systems analyze this data to predict embryo
approach to the problem in the future. 3,4 viability. Selected embryos are then processed using
AI-based ART. Pregnancy outcomes are tracked, enabling
ART techniques, such as intracytoplasmic sperm
injection (ICSI) and in vitro fertilization (IVF), are continuous refinement and optimization of the AI systems.
frequently utilized to assist infertile couples in getting Ethical considerations and regulatory compliance are
pregnant. The procedures involve the retrieval of eggs, crucial at every stage of this process. This systematic review
laboratory fertilization with sperm, transfer of viable explores the following research queries (RQs):
embryos into the uterus, and control of ovarian stimulation. • RQ1. What are the current state-of-the-art AI
The viability of ART is highly dependent on the quality of technologies used in embryo selection for ART?
gametes and embryos, which are conventionally evaluated • RQ2. To what extent does ART improve the pregnancy
subjectively by embryologists based on morphological success rate and live birth outcomes compared to
traditional methods?
criteria.
• RQ3. How can AI algorithms be seamlessly integrated
Since the birth of Louise Brown, the first infant conceived into existing embryo selection protocols and
through IVF, ART has undergone significant advancements laboratory workflows to leverage the expertise of
aimed at reducing complications and improving embryologists and healthcare professionals?
outcomes. The integration of artificial intelligence (AI) • RQ4. What is the impact of AI-driven embryo
5
into ART holds great promise for enhancing outcomes. selection on the psychological health and decision-
AI technologies, such as computer-assisted sperm analysis making processes of prospective parents, particularly
and machine learning (ML) algorithms, enable the in light of ethical concerns?
objective evaluation of semen parameters and embryo • RQ5. What are the primary barriers and limitations
quality. By standardizing evaluations and processing large to the clinical application of AI algorithms in embryo
volumes of data, AI has the potential to enhance treatment selection, and how are these technologies being
outcomes and increase conception rates. However, developed and validated?
6-8
despite its groundbreaking breakthroughs, only 30% of
ART treatments result in conception, highlighting the 2. Research methodology
necessity for more accurate predictive models. As the This section outlines the research design and analytical
9
process involves manipulating human gametes or embryos procedures used in the present study.
in vitro, ART outcomes are influenced by multiple
complex factors, including the cause of infertility, age, 2.1. Overview
hormonal profile, and laboratory conditions. Advanced According to McKenzie et al., a systematic review is a
15
technologies, such as AI and ML, are being explored to rigorous, structured method for identifying, evaluating,
enhance prediction accuracy and decision-making in and synthesizing research evidence on a specific
ART, with promising results in predicting IVF cycle question using predefined protocols. It minimizes
10
outcomes and guiding embryo selection. These systems bias through comprehensive literature searches and
11
interpret data using image-based analysis to provide transparent processes, often incorporating meta-analysis
clinically relevant recommendations, and AI models are to quantitatively combine study findings. Consequently,
12
also being developed to classify reproductive data, such as this systematic review was conducted to investigate the
embryonic development and semen characteristics. As 4,11- technologies utilized in AI-guided embryo selection and to
13
13 AI continues to demonstrate potential in improving map the current landscape of advancements within ART.
diagnostic and therapeutic processes in reproductive
medicine, its adoption in fertility clinics is likely to 2.2. Objectives
increase. Nonetheless, challenges remain regarding the The key objectives of this systematic review include:
14
generalizability and standardization of AI applications in (i) To examine current AI applications in embryo
ART. 4 selection and analyze the success rates of various ML
This systematic review aims to identify and map models used in ART
the current landscape of research on embryo selection, (ii) To identify potential future improvements,
focusing on advancements and innovations in AI-based innovations, and existing research gaps in the
ARTs. In Figure 1, the flow diagram depicts the integration application of AI for embryo selection

Volume 2 Issue 3 (2025) 2 https://doi.org/10.36922/aih.7170