Global Health Asia-Pacific Issue 1 | 2023 GHT64B | Page 58

AI Fertility
I believe AI will be a game changer in the future , not just on embryo selection and genetic screening , but in fertility medicine in general because it will help us get at the right answers in a faster and safer way , ultimately ensuring patients a shorter and easier journey towards a successful pregnancy . laboratory technologies , the success rates for IVF have remained stagnant without noticeable changes for the past decade . Part of the reason is because every patient has different conditions and there ’ s no one treatment that works for all . The planning of IVF treatment regime relies heavily on a clinician ’ s input . There are a lot of decision makings throughout the entire treatment and they are often affected by clinician ’ s experience , difference in training and in house practice . That ’ s where AI can come in , as it can reduce those variables by offering a reliable system built from huge size of evidence-based data collected worldwide . In short , it can offer the best treatment option with better accuracy and consistency .
AI also has the potential of taking over repetitive and routine tasks such as managing medical record , health monitoring , etc . This will help to reduce financial cost and time lost due to human error and practice variations .
I believe AI will be a game changer in the future , not just on embryo selection and genetic screening , but in fertility medicine in general because it will help us get at the right answers in a faster and safer way , ultimately ensuring patients a shorter and easier journey towards a successful pregnancy .
Do you think the use of AI in fertility medicine should be an important component of an embryologist ’ s training ? Tee Sze Tian : Yes , upcoming scientists should be trained more on how to operate the AI system as AI will inevitably become an integrated part of fertility medicine . The benefits of AI to improve the productivity and efficacy in the IVF lab can only be maximised when the operators are very familiar with its functions .
Reports highlight that Malaysia currently has a limited number of certified embryologists . Do you think the greater use of AI will reduce the demand for embryologists in the country ? Tee Sze Tian : I do not anticipate the demand for experienced embryologists to decrease for the next twenty years , as embryology is a highly skill dependent role . The most important tasks of the embryologists are being performed at microscopic level and turning them into AI-driven automated process is still a challenge at the moment . A significant portion of an embryologist ’ s role also involves dealing directly with patients to understand more about their pathology , medical history , and mental condition . In addition , the AI system requires human participation to input patient information accurately and comprehensively . Any error in the data input will risk incorrect prediction generated by the system .
Is Malaysia ’ s healthcare sector , particularly in fertility medicine , ready for widespread use of AI ? Tee Sze Tian : The fertility professionals in Malaysia always keep themselves updated with the latest advancements in assisted reproductive technology . We are dedicated to ensuring Malaysian couples who need fertility assistance have access to the latest technologies and treatment options , like what TMC Fertility did with the application of AI . A challenge we may face with the use of AI in Malaysia is the limitation of the population size . As the accuracy of AI relies on the dataset it was trained with , the algorithms currently available may be limited to the population or community which the platform was designed for . We may need a little longer to establish algorithms that would work best for Malaysians as we have a smaller population size and we need more time to acquire a good size dataset .
How is AI applied in fertility medicine and which problems can the technology help to solve ? Adelle Lim : Embryologists are still using morphology to assess embryos , meaning that they look at them and select those that look nice . But this is a very subjective and sometimes biased process as different embryologists can provide different scores for the embryos involved .
In contrast , AI is more objective and unbiased in providing a score for the embryo image from 1 to 10 , with a higher score indicating higher chances of a successful pregnancy .
How often do embryos produced through IVF develop genetic abnormalities and which are the most common ? Adelle Lim : We cannot deny that embryos grown in the lab have a high incidence of abnormalities . For example , healthy young patients who undergo IVF will have a 30 percent chance of getting abnormal
AI can help select the best embryos fertilised in the lab
56 ISSUE 1 | 2023 GlobalHealthAsiaPacific . com