Media
HKU Faculty of Dentistry launches world’s first AI-powered clinic for early oral cancer detection
11 May 2025
The Division of Oral and Maxillofacial Surgery (OMFS) of the Faculty of Dentistry at the University of Hong Kong (HKU) and the Prince Philip Dental Hospital (PPDH) have jointly launched an AI clinic dedicated to patients with oral potentially malignant disorders (OPMD). This initiative aims to advance the prevention and early detection of oral cancer. As the world’s first of its kind, this AI clinic commenced as a new medical service at PPDH in April 2025.
Oral cancer is a malignancy that affects the lining of the lips and mouth. It is the most common tumour in the head and neck region. Globally, the survival rate of oral cancer after five years of diagnosis is about 50%, mainly because many patients are diagnosed at a late stage. This represents approximately 60% of all oral cancer patients in Hong Kong. As a result, there is a need to intensify efforts and public action to promote city-wide oral cancer prevention and early detection.
Oral cancer has the potential to be detected early because long before cancer develops, some initial lesions may appear in the mouth in the form of white or red plaques, known as oral potentially malignant disorders (OPMD). White plaques such as oral leukoplakia (OL) and oral lichenoid diseases (OLD) are the most common subtypes of OPMD. Since individuals with OPMD have a significant higher risk of developing oral cancer compared to the general population, timely detection and treatment are crucial for oral cancer prevention and early detection.
The AI clinic is equipped with an AI-based web tool called ‘OralCancerPredict’, developed by the research team at the HKU Faculty of Dentistry in 2021. This tool serves as a decision-support system to guide clinical treatment recommendations for patients and promote cancer prevention and early detection. The tool is based on an AI model trained using a large cohort of patients with OL and OLD treated at the Queen Mary Hospital between 2003 to 2022 and validated on similar patient cohorts in the UK and Nigeria.
The tool utilises multidimensional information, including patient demographics, comorbidities, clinical appearance, and histological features, to calculate the overall probabilities of cancer developing in OL and OLD. It also provides risk classifications (high-risk and low-risk status) and explains the rationale behind the predicted risks. Surgery is recommended for high risk patients, while observation, counselling, and supportive therapies are recommended for those with low risk.
Previous studies have demonstrated that OralCancerPredict achieves an average accuracy of 94%, exhibits good calibration, and offers superior potential clinical benefit compared to current methods used to assess cancer risk of OL and OLD in clinical practice. This new technology provides enhanced support to doctors in identifying patients at the highest risk of oral cancer, enabling early treatment and close follow-up. Moreover, using the AI web tool to streamline patient management can improve clinical service delivery by avoiding unnecessary surgery, particularly for low-risk patients, while maintaining regular disease surveillance every six months.
The AI clinic is led by Professor Richard Su, Chief and Clinical Professor of the OMFS, with contributions from Clinical Associate Professor Liwu Zheng, Clinical Assistant Professor Jane Jingya Pu, Dr Joanna Ko from the OMFS, and Assistant Professor John Adeoye from Digital/Precision Dentistry.
The clinic is open to all patients with OL and OLD, who will receive treatment and periodic follow-up as part of the medical service. Incorporating the AI web tool to support treatment decisions will not increase treatment costs, as the AI models are free of charge and do not require additional tests. Patients wishing to join the medical service must provide a referral from a general dental or medical practitioner.
To learn more about OralCancerPredict, please visit: http://oralcancerpredict.facdent.hku.hk
For media enquiries, please contact:
Professor John Adeoye
Assistant Professor in Digital / Precision Dentistry
Applied Oral Sciences and Community Dental Care, HKU Faculty of Dentistry
Email: [email protected]
Tel: (852) 2859 0215