Joseph Wu
Professor, University of Hong Kong
Thu
,
Jul 30, 2026
11:00 am
EST
Jul 30, 2026
11:00 am
In-person
Portsoken Street
London, E1 8PH, UK
London, E1 8PH, UK
The Roux Institute
Room
100 Fore Street
Portland, ME 04101
Portland, ME 04101
Network Science Institute
2nd floor
2nd floor
Network Science Institute
11th floor
11th floor
177 Huntington Ave
Boston, MA 02115
Boston, MA 02115
Network Science Institute
2nd floor
2nd floor
Rm 140, 3rd Fl, 101 Blvd
Talk recording
This talk comprises two applications of advanced data science methodologies for optimizing clinical workflows and public health screening strategies for cancer.
First, we will discuss the development of an AI-assisted oncology natural language processing (NLP) pipeline. We utilize large language models (LLMs) to automatically extract critical cancer staging and risk stratification data from unstructured clinical notes, with a focus on thyroid cancer and hepatocellular carcinoma in Hong Kong. Using open-source LLMs, the framework achieves over 90% accuracy in American Joint Committee on Cancer (AJCC) staging and reduces clinicians' pre-consultation preparation time by approximately 50% through the use of an automated clinical dashboard.
Second, we develop a natural history model that characterizes the epidemiology of colorectal cancer (CRC) in Hong Kong. By leveraging quantitative fecal immunochemical test (FIT) data from over 248,000 participants in the Hong Kong CRC Screening Program, we infer age- and gender-specific prevalence of colorectal neoplasms and annual stage progression rates which are very difficult to estimate directly from empirical studies.
Together, these projects demonstrate how leveraging diverse data sources. from free-text clinical records to population-level screening data, can significantly advance oncological care and precision screening interventions.
About the speaker
Joseph Wu is the Sir Robert Kotewall Professor in Public Health at The University of Hong Kong (HKU). He specialises in the mathematical and statistical modelling of diseases, aiming to develop practical analytics and strategies for disease control and prevention. His current research focuses on AI tools for global and personal health protection, particularly in clinical natural language processing, epidemic nowcasting and forecasting, and vaccine hesitancy. Professor Wu has worked on COVID-19, seasonal and pandemic influenza, hand, foot and mouth disease, HPV, MERS, yellow fever, cervical cancer, colorectal cancer, and breast cancer. He earned his PhD in Operations Research and his BS in Chemical Engineering from MIT. He is the workstream modeling lead at the Hong Kong Jockey Club Global Health Institute, a tripartite collaboration between The University of Hong Kong, the University of Cambridge, and the International Vaccine Institute. Professor Wu directed HKU’s first Massive Open Online Course, Epidemics, which has enrolled more than 50,000 learners since its launch in 2014. He is the director of two Croucher Summer Courses, on Vaccinology and Vaccine Hesitancy; Co-Editor-in-Chief of Epidemics; and an Associate Editor of PLOS Computational Biology and PLOS Neglected Tropical Diseases. He is a Fellow of the UK Faculty of Public Health and served as a member of the WHO Advisory Committee on Immunization and Vaccine-related Implementation Research from 2018 to 2023.
Share this page:
