Dr Hrishikesh Pai, Founder & Medical Director, Bloom IVF
The field of IVF has advanced drastically since the first IVF baby was born in 1978. In the beginning, the success rates were low in the single digits, but with advances in reproductive sciences and laboratory techniques, the success rate in IVF has increased to around 30 per cent. Today, we are witnessing an increasing incidence of infertility from various factors including lifestyle changes and rise in diseases such as obesity, diabetes and PCOS (polycystic ovary syndrome). This is resulting in a growing demand for IVF and a need to reduce the unpredictability and number of attempts for couples finding it difficult to conceive. It is here by that Artificial Intelligence (AI) can play a major role.
AI in IVF
AI has transformed the practice of medicine since it was first used in the 1970s to help doctors identify the right treatment for blood infections through an AI program called MYCIN. Decades since then have seen AI evolving rapidly to impact various aspects of healthcare delivery from data processing to diagnosis and treatment across specialities to make medicine more personalised. The field of fertility medicine is also seeing AI making a significant contribution to help specialists navigate the complex array of factors that come into play when considering in-vitro fertilization (IVF) and improve the success rate of the process that is otherwise lengthy and emotionally challenging for couples who want to be parents.
AI can contribute at various stages of the IVF process that combines medical and surgical procedures. It can predict the oocyte potential, assess sperm and egg quality, and determine the viability of embryos. AI can help in reducing the number of embryos transferred, lower the risk of multiple pregnancies and reduce the time taken to become pregnant. It will also help in reducing the number of patients who discontinue IVF treatment due to physical and psychological strain.
Assessment of female reproductive function
AI in ultrasound is used for imaging of the ovaries to monitor follicles. It can also assist in the evaluation of the uterus and assess endometrial readiness and receptivity. AI can predict the implantation potential based on the condition of the uterus and endometrium. This helps in predicting the outcome of IVF and embryo transfer.
Planning for IVF cycle
Conventional planning for an IVF cycle depends on the age of the patient and their medical history. Specialists may specify treatment regimens based on their clinical experience. An IVF cycle might thus vary in the protocol for stimulation and mode of fertilization as well as the use of other procedures such as assisted hatching and preimplantation genetic testing. AI can support reproductive specialists by enabling objective decision-making to optimize the therapy regimen. AI algorithms can also help in the selection of the ideal hormone-replacement treatment for patients requiring it to increase their chance of successful implantation.
Precision in embryo selection
The process of IVF includes assessments at various stages that are done by fertility specialists. One of the critical aspects is selecting the embryo for implantation. The conventional method involves a visual evaluation of the quality of an embryo by an embryologist to choose one that they think is most likely to result in a healthy pregnancy. Embryologists use methods such as microscopy and time-lapse monitoring to determine whether an embryo, which at that stage consists of just 200 to 300 cells is viable or not. This brings in an element of subjectivity as embryologists have different systems of categorizing embryos. The experience of the embryologist is also another factor.
AI helps in changing this by bringing objectivity and reliability to the selection process by helping the embryologist identify the embryo that will maximise success. AI standardises the process of embryo selection with automation making it more precise by reducing bias and the dependence on subjective human judgement. AI is trained to improve and automate the process of embryo ranking and selection based on the potential to implant through algorithms based on relevant information from a large number of embryo microscopy images. Today, AI products that are proven in clinical studies are available for uniform and accurate assessment of embryo quality.
The future
IVF is a process involving sequential decisions and AI technology can help in decisions across the decision chain with analysis of all embryological, clinical, and genetic data to provide personalised fertility treatments. The future will see AI empowering fertility specialists to make smarter data-driven decisions. AI technologies will go beyond the current limited use for embryo selection to give us new insights into causes of infertility from patterns in patient data and usher in a new era of standardisation, automation and precision in fertility medicine.