Artificial intelligence in reproductive medicine

Abstract

Artificial intelligence (AI) has experienced rapid growth over the past few years, moving from the experimental to the implementation phase in various fields, including medicine. Advances in learning algorithms and theories, the availability of large datasets and improvements in computing power have contributed to breakthroughs in current AI applications. Machine learning (ML), a subset of AI, allows computers to detect patterns from large complex datasets automatically and uses these patterns to make predictions. AI is proving to be increasingly applicable to healthcare, and multiple machine learning techniques have been used to improve the performance of assisted reproductive technology (ART). Despite various challenges, the integration of AI and reproductive medicine is bound to give an essential direction to medical development in the future. In this review, we discuss the basic aspects of AI and machine learning, and we address the applications, potential limitations and challenges of AI. We also highlight the prospects and future directions in the context of reproductive medicine.

Figure 1

The role of artificial intelligence in Reproductive Medicine. Big data include electronic medical records (EMRs) and other data. EMRs can capture data from various ways and the data is analyzed using AI such as machine learning and natural language processing (NLP). AI has been used in the many aspects of reproduction, from research and experiment to clinical practice. This schematic reviews the seven main applications of AI in reproductive medicine.

Figure 2

The workflow of artificial intelligence in Reproductive Medicine. This flowchart provides a brief overview of the AI workflow. The first step is the collection of data. The data includes electronic medical records (EMRs), hospital data and cloud data sharing. The second step is data pre-processing. The third step is the selection of the appropriate model. The data is analyzed using artificial intelligence methods such as machine learning and natural language processing (NLP). Then the training dataset is used to train the model. The final steps include the evaluation and validation of the model.

Figure 3

AI applications in reproductive medicine. (A) This decision tree model is conducted to make predictions for the selection of embryos. The model first separates cumulus cells samples upon AMHR2 expression (high or low) and then upon LIF expression (high or low). The gray color represents high-quality embryos and the white color represents low-quality embryos. The combination of high AMHR2 and low LIF expression achieves an 82.6% possibility of predicting a low-quality embryo, and the combination of low AMHR2 and high LIF expression leads to a 74.6% possibility of predicting a high-quality embryo (Devjak et al. 2016). (B) The researchers obtained human semen samples from eight healthy donors and acquired the quantitative phase maps of the sperm samples by using the diagram of the optical system. Then they used a program to extract the phase map and features. Finally, the dataset obtained was used to train a two‐class SVM classifier (Mirsky et al. 2017). (C) (Girela et al. 2013) created an ANN model to produce a decision support system that can help predict the semen parameters based on the data collected by the questionnaires and can support the traditional diagnosis.