The 4th session of the “This is Global Health!” Symposium Series (Fall Semester), hosted by the Global Health Research Center at Duke Kunshan University, concluded successfully on November 21. Centered on the theme “How to Prepare for the AI Era in Global Health?”, the event featured a keynote address by Professor Tien Yin Wong—Professor & Senior Vice-Chancellor of Tsinghua Medicine and Vice-Provost of Tsinghua University (Beijing, China), and an international authority in ophthalmology. Offline attendees included faculty, students, and members of the Kunshan Health Commission Delegation, while a global online audience also joined this interdisciplinary dialog to explore the integration of artificial intelligence (AI) with global health.
In his keynote speech, Professor Tien Yin Wong, drawing on his extensive interdisciplinary practical experience, systematically analyzed the core challenges in the field of global health, practical obstacles to AI implementation, and pathways for building an AI-empowered global health ecosystem.
He stated that the global health sector is currently undergoing profound changes and facing multiple challenges: the changing disease spectrum, marked by aging and rising chronic disease prevalence; significant disparity in access to high-quality healthcare, with primary care systems particularly underdeveloped in low- and middle-income countries (LMIC); escalating healthcare costs; and shortages in both the quantity and expertise of healthcare workers.
Although the industry regards breakthroughs and scaled-up applications of AI technology as potential solutions to the above trends and challenges, the actual implementation process has been unsatisfactory. For instance, insufficient understanding of local healthcare ecosystems in different regions leads to poor technical adaptability; lack of validation in real-world clinical scenarios for AI models, undermining their generalizability; absence of guidelines for developing and evaluating medical AI systems; low levels of trust and engagement with AI among clinicians and patients; acute shortages of data and studies in LMIC; and the challenge of aligning AI applications with patient-centered, humanistic care.
Professor Wong shared the practical experience of Singapore’s “Selena Plus” project—the world’s first national AI-driven screening program for diabetic retinopathy. This project overcame the limitations of traditional screening which relied heavily on specialists, translating AI innovation from research labs into tangible public health impact and protecting the vision of tens of thousands of individuals. Building on practical experience from this and other projects, he proposed the “Six Ps” framework to guide the development of an AI-enabled global health ecosystem:
· People: Cultivate interdisciplinary talent pipelines, transform medical education, and train AI-ready physicians;
· Product: Develop trustworthy AI tool that are explainable, interpretable, and ethically sound, shifting from “Disease-centric” models to those focused on “Universal Health”;
· Platform: Establish secure data-sharing platforms, balancing privacy protection with data utilization through approaches like synthetic data generation and differential privacy;
· Policy: Develop unified guidelines for medical AI model development and evaluation;
· Process: Streamline workflows for integrating AI into healthcare systems and build an evidence base of real-world applications;
· Partnership: Advance public-private partnerships and international collaboration to address data bias and advance global health equity.
Against the backdrop of this technological transformation, Professor Wong also shared Tsinghua Medicine’s development strategies: building an integrated academic medical system, promoting the translation of medical technologies, and cultivating clinical medicine and medical research talents with interdisciplinary backgrounds who can adapt to future technological advancements.
During the roundtable discussion, Professor Tien Yin Wong and Shixin Xu, Assistant Professor of Mathematics at Duke Kunshan University, addressed and discussed questions raised by faculty and students. Regarding the question of “how physicians can balance the use of AI tools and the retention of core clinical skills”, Professor Wong took gastroscopy biopsy as an example, emphasizing that AI can only indicate suspicious lesion areas, while the final diagnosis must be made by physicians based on their clinical experience. He stressed that medical education should incorporate courses on AI ethics and skill maintenance to avoid the degradation of core diagnostic and treatment capabilities.
In response to the question of “whether involving physicians in mathematical and AI model training can better facilitate the integration of AI into the healthcare ecosystem”, Assistant Professor Shixin Xu pointed out that pure data-driven AI faces significant challenges in data acquisition. This can be addressed by generating synthetic data through mathematical modeling and calibrating it with real-world data. He also suggested that physicians should focus on standardizing data at the source rather than learning complex mathematical knowledge.
This symposium clearly outlined the development blueprint for global health in the AI era, providing forward-looking and practical references for faculty, students, and healthcare practitioners.
Written by Ruoning Feng
