This article first appeared in Digital Edge, The Edge Malaysia Weekly on March 15, 2021 - March 21, 2021
In vitro fertilisation (IVF) is a daunting, not to mention costly, process for parents desperately wishing to conceive. However, the success rate of the treatment is still, on average, under 50%.
Practitioners at the Alpha IVF and Women’s Specialists Centre Kuala Lumpur are, however, working towards improving the odds of conception, with the use of artificial intelligence (AI).
By employing AI in a technique developed to improve embryo selection, the chances of a successful pregnancy have increased, says Datuk Dr Colin Lee, the centre’s medical director.
“Matching sperm with an egg, selecting an embryo and then implanting it to produce a full-term pregnancy is filled with variables and some patients may need to undergo two or more cycles before they are successful.
“Even if a prospective couple goes to a Malaysian fertility centre for IVF, no one can guarantee that pregnancy will happen immediately. We do our best to raise the chances, but nature always deals in probabilities, and never certainties,” says Lee.
Alpha IVF and Women’s Specialists Centre is one of the premier conception clinics in the country and has a success rate of up to 82.9% when it comes to IVF. This is because Lee has been at the forefront of using precision technology in his medical practice.
According to the Reproductive BioMedicine Online journal, 2.5 million IVF cycles take place every year, resulting in over 500,000 deliveries. Malaysia has a very high success rate that is comparable with that of some of the most advanced countries in the world.
Now, Lee hopes to further minimise the occurrence of failed pregnancies by working with Life Whisperer — an Australian start-up that has developed a technology to identify embryos far more successfully than humans by analysing images of them and comparing these with pictures in a vast database.
Using machine learning, the technology identifies traits of embryos from around the world that had previously gone on to result in successful pregnancies via IVF.
“Recent studies have demonstrated a 32% improvement in accuracy when it comes to embryo viability assessment compared to the traditional embryologist morphology assessment,” says Lee.
AI has been used in precision medicine for years, especially in learning how to predict clinical outcomes of specific interventions or procedures, thereby improving the outcome.
Alpha IVF is among the eight centres that have been part of Life Whisperer’s embryo selection software study since 2019 to improve pregnancy outcomes using embryo images. The company started commercialising the software last year.
“By using these tools, we are able to predict which embryos are likely to result in a viable pregnancy. We are the first in the country to use such a technology,” says Lee.
Of the eight centres in the study, Alpha IVF is the only one that uses embryoscope — a state-of-the-art technology that enables a stable culture environment for embryos, he adds.
“Through embryoscope, we analyse and study the embryo as it grows. Throughout the five days of growth, the software takes time-lapse photos at 10-minute intervals. There is no flash; we do not want to disturb the embryos. The photos become a video.
“There is now immense data on the development of each embryo compared to what is normally done — where our scientists have to check the embryo two or three times during its growth until it reaches blastocyst. This means they check in the beginning, middle and end. Whereas with AI, the embryo is photographed every 10 minutes,” says Lee.
Blastocyst refers to the stage that the human embryo reaches five to six days after fertilisation. The embryo must reach this stage before it can hatch and implant in the lining of the uterus.
“IVF is not a procedure, or even a process. It is a programme that lasts for a few months before the final act, which is the embryo transfer. So, an IVF programme is a long chain of processes, methodologies, techniques and research and development (R&D). It is important to remember that a chain is no stronger than its weakest link.
“Let’s say a chain has 200 links; 199 links are made of steel, but one link is made of thread. It’s going to give way at the thread, no matter how strong the other links are.
“What we have done over the years is [we have] identified every single link and fortified the weak ones. Whatever is made of thread, we turn into gold so the whole chain becomes very strong,” he says.
According to Adelle Lim, the centre’s chief embryologist, the AI has been fed tens of thousands of images in the last couple of years and continues to learn as centres around the world continue to nourish the programme.
“It is a web-based application that is completely non-invasive. Our embryologists can just drag and drop images of day-five embryos and the AI provides a score for each one relating to its implantation potential and likelihood of creating a pregnancy.
“The embryo images contain the outcome of each embryo — whether the embryo is implanted or not implanted. The programme then develops the algorithm to scan the embryo images that we put in and gives the embryo a score up to 10. The higher the score, the higher the chance of implantation,” says Lim.
So far, the AI software has been able to provide 25% more accurate assessments than an experienced embryologist, adds Lee.
“AI uses objective data, whereas with an embryologist, there’s a lot of subjectivity involved. If the embryologist had a bad day or even a bad night’s sleep, it may affect their assessment of the embryo.”
The technology was approved to be rolled out in Malaysia last November and Alpha IVF is one of the first to use AI that ensures the health and survival of embryos, says Lee.
“This technology will add to our arsenal of techniques that will hopefully save our patients’ time and money, and lead to successful pregnancies.”
The cost of an IVF treatment generally starts at RM20,000. “However, things like frozen embryo transfer or checking the chromosomes of a blastocyst will add over RM10,000 to the bill,” says Lee.
And while there is no guarantee of a successful pregnancy, the technology does provide a definite advantage, he adds.
Having AI improve the IVF process is expected to be a boon for many individuals striving to conceive late in life. “If left to nature, the optimum age to bear a child would be in one’s late teens. Infertility becomes worse as they age, notwithstanding the impact of the environment.
“Generally, up to one’s mid-30s, one is quite fertile. But in the late 30s, things get more challenging, and in the early 40s, there is a huge drop in the ability to get pregnant and the ability to produce a healthy child.
“Most women don’t know that they don’t produce any new eggs in their life. All their eggs are produced when they are foetuses. One foetus produces about six million eggs and by the time the child is born, there is only about one million left.
“By the time they reach puberty, there are only 250,000 to 300,000 eggs left. After that, they progressively lose these eggs as time goes by, and by the time one reaches menopause, there are only 1,000 eggs left. And then there is the poor quality of the eggs at later stages in life that makes the process of conceiving naturally even more difficult,” he explains.
The continued use of current technology is why the clinic has one of the highest pregnancy rates in the world, says Lee.
“We have painstakingly identified any area that is weak or sub-optimal and this is where we have been doing R&D for years. In the last two decades, we have produced about 200 scientific papers.
“We are looking at other applications of AI to achieve the ultimate goal of patients coming to see us, which is to have a baby. And, not just a baby but a healthy baby,” says Lee.
Save by subscribing to us for your print and/or digital copy.
P/S: The Edge is also available on Apple's AppStore and Androids' Google Play.