An interview with Dr Bich-Tram Huynh
Insights into the quest for diagnostic accuracy through AI
1. What is your background and what led you to pursue this line of research?
I’m currently a research director at the Institut Pasteur, where I lead a group focusing on maternal and child health in low- and middle-income countries. I earned my MD from Paris Descartes University in 2006 and completed my PhD in Epidemiology from Pierre and Marie Curie University in 2011. Over the last 16 years, I’ve worked in diverse settings such as Benin, Togo, Madagascar, Senegal, and Cambodia, which has given me a wealth of experience in field epidemiology, public health, and biostatistics. These experiences have profoundly shaped my research, driving me to focus on developing innovative solutions to improve health outcomes for women and children in resource-limited settings.
2. Can you give a brief overview of the project you are currently supervising?
This [GIFT] project focuses on developing a risk assessment screening tool for reproductive tract infections (RTIs), including sexually transmitted infections (STIs) and bacterial vaginosis (BV). The objective is to improve on the limitations of the 2003 and 2021 WHO syndromic management guidelines, which often fail to identify asymptomatic cases. We aim to create a simple, cost-effective tool that can be applied across various populations and demographics to improve early detection and treatment of RTIs.
3. What is the primary aim of this project?
The primary aim is to design a risk assessment tool that effectively screens for STIs and BV, even before women undergo clinical testing. By identifying high-risk women – especially those without symptoms – we hope to increase the number of women who seek testing and treatment, thereby reducing the overall burden of untreated infections and improving women’s reproductive health outcomes.
4. How has AI, particularly machine learning, changed the way we approach the development of health-related algorithms and diagnostics?
AI and machine learning have transformed how we approach health-related algorithms by enabling us to process large datasets efficiently and identify patterns that might be missed using traditional methods. This advancement is particularly impactful for diagnostics as it enables:
Tailored Approaches: AI allows us to personalise diagnostics and treatments to individual patient profiles.
Predictive Capabilities: Machine learning models can predict risks by analysing both historical and real-time data.
Scalability: These models can be scaled to accommodate large, diverse datasets, which enhances the accuracy and applicability of diagnostic tools across different populations.
5. What excites you most about the potential impact of this research?
I’m mostly excited about the potential to make a tangible difference in the lives of women in low- and middle-income countries. By developing a tool that can accurately detect STIs in settings where resources are scarce, we have the opportunity to significantly improve health outcomes, reduce the burden of disease, and enhance overall quality of life.
6. What has been the most rewarding part of leading this project?
The most rewarding part has been seeing the real progress we’ve made toward developing a practical tool that addresses such a critical healthcare need. Collaborating with a multidisciplinary team, witnessing the application of cutting-edge technology to real-world problems, and knowing that our work could have a profound impact on global health are incredibly fulfilling.
7. What advice would you give to researchers or students interested in women’s health and STIs?
I would advise researchers and students to embrace an interdisciplinary approach, as it greatly enhances the ability to solve complex health challenges. Also be sure to maintain a sense of curiosity and persistence throughout your research journey.