No child's play: How AI is helping medicine in assisted fertility

The technology helps doctors by studying massive amounts of data. It's useful but involves ethical concerns

Divya Nayar, 30, a senior manager in a technology company in Bengaluru, is a fan of science fiction. Adithya, her husband and colleague, is not. When the couple’s obstetrician-gynaecologist (OB-GYN) suggested that artificial intelligence (AI) “guide” their fertility procedure, Divya said yes. Adithya, 32, was sceptical. “So, we are having robot babies now?” he recalled asking. (Their names have been changed for privacy.)

Amid rising excitement about AI, science-fiction scenarios seem within the reach of realisation. ChatGPT, the popular AI chatbot that responds to a range of written queries, may have already turned The Great Automatic Grammatizator, Roald Dahl’s short story about a machine writing a prize-winning novel in minutes, into a reality.

In the 2019 Australian cyberpunk thriller I Am Mother, a robot grows a human embryo and cares for her. AI, which is being used to detect cancer, pinpoint microscopic cavities, and answer complex medical questions, is assisting fertility doctors select the ideal embryo for in-vitro fertilisation (IVF). For couples like the Nayars, a sci-fi narrative has begun to take root in reality.

Health start-ups in the United States, Europe, and West Asia have made progress in using AI for fertility treatment. AIVF, a reproductive technology company based in Tel Aviv in Israel, has developed an AI-powered software called EMA that processes vast amounts of data — beyond what a human can manage — to assist the embryo-selection process.

Medical tool

IVF clinics and health-tech start-ups in India are proposing AI-aided fertility services to young couples like the Nayars. Santaan, an Odisha-based start-up that offers various fertility services, uses AI in selecting embryos for transfer to the womb.

Such selection is crucial for the success of an IVF cycle, as only a small number of embryos are transferred to wombs. AI algorithms can analyse images of embryos and predict which are most likely to lead to a successful pregnancy. AI can help reduce the number of unnecessary embryo transfers and minimise the risk of multiple pregnancies.

Santaan isn’t the only one, of course. Last year, Delhi-based Indira IVF tied up with AI-healthcare firm Presagen to roll out an internationally tested AI tool called ‘Life Whisperer’ in more than 100 clinics in India. This tool too improves embryo selection.

Dr Nitiz Murdia, co-founder and managing director of Indira IVF, explained the process of embryo selection through AI. “Traditionally, embryologists visually assessed embryos for viability and likelihood of clinical pregnancy. Now, AI-powered software, trained on thousands of embryo images, evaluates embryos with greater speed and precision, providing standardised grading,” he said.

Embryo selection is not the only aspect of fertility care that AI is helping with. Machine learning (ML) is used to select oocytes — female germ cells involved in reproduction — and assess their behaviour during intracytoplasmic sperm injection (ICSI), an assisted reproductive technology (ART) procedure.

AI methods have shown promise in analysing embryo development and implantation potential. They are used for semen analysis and DNA integrity, assisting embryologists to select sperm cells in male patients suffering from infertility.

AI can help optimise cryopreservation techniques – used to keep live cells, tissues and other biological samples in a deep freeze – by analysing large datasets of frozen embryo outcomes. By identifying patterns and factors that influence the viability of thawed embryos, AI can then aid in developing improved protocols and strategies for preserving embryos during the freezing and thawing processes, said Ashootosh Chand, partner for digital and emerging technologies at PwC India.

AI also helps to better assess the uterus and ovaries and identify uterine anomalies.

“Our OB-GYN has also assured us that the AI assistance will increase the success rate of the IVF procedure,” said Divya Nayar.

Ethics of it

Murdia said that while AI technology improves the embryo selection process, it is important to note that pregnancy success rates depend on individual patients. “We have witnessed improved clinical workflow, reduced manual evaluation time, and an increased potential for successful pregnancies,” he said.

According to Murdia, estimates suggest 12 million people have been born through IVF – the population of Bengaluru almost – in the last 45 years in India. As many as 280,000 IVF treatments were done in India in 2022. “With the advent of technology, especially the revolutionising prowess of AI, these numbers are further slated to increase” he said.

Dr Vaishali Chaudhary, director-IVF & fertility at Sahyadri Hospitals in Pune, said implementing AI in ART has challenges. A major one is the need for large amounts of precise data. 

While AI algorithms can provide quick and objective assessments, it is essential to validate their predictions. This involves comparing AI-generated selections with clinical outcomes and continually refining algorithms to improve accuracy. Fertility clinics using AI tools will have to overhaul their infrastructure and they will need embryologists who can adapt to the technology’s limitations.

Ethical issues and patient privacy is the main challenge. “ART processes involve highly sensitive personal information, and AI technologies must adhere to strict data protection protocols to maintain patient confidentiality and prevent unauthorized access” said Murdia of Indira IVF.

The Nayars said that opting for AI-aided IVF has meant additional medical exams and more rigorous and repeated questionnaires about their families' health history.

Dr Satish Prasad Rath, founder of Santaan Group of Institutions in Odisha, said AI-driven ART processes are expensive. Santaan is collaborating with the state-run Indian Institute of Technology Hyderabad to develop an indigenous and affordable solution in this field.

Anubhav Prashant, COO at Hyderabad-based Apollo Fertility, said changes are needed for more successful adoption of AI in ART. “Data quality, regulatory compliance, ethical considerations, user trust, and the need for continuous improvement and validation are some of the challenges we face every day. Addressing these challenges requires a comprehensive approach involving collaboration with medical professionals, regulatory bodies, data experts, and user feedback,” he said.

Mayank Makkar, Founder and CEO at Rivi AI concludes, "AI should augment, not replace, human decision-making in fertility care."