New Faculty AI Clinic supports oral cancer detection and clinical decision-making
The AI clinic for patients with OPMD where the AI-based methods will be used as decision-support tools to guide clinical treatment recommendations for patients.
A team at the HKU Faculty of Dentistry, led by Professor Richard Su, Clinical Professor in Oral and Maxillofacial Surgery, has partnered with the Prince Philip Dental Hospital to launch an Artificial Intelligence (AI) Clinic that is dedicated to helping patients at risk of oral cancer.
“Launched in April 2025, our AI Clinic is the first of its kind worldwide to use AI to promote the prevention and early detection of oral cancer,” said Professor Su. “Our AI-powered medical service assesses if oral cancer risk among patients is high or low, and then guides dentists and doctors to take appropriate action.”
The clinic relies on an AI-based predictive assessment tool called “OralCancerPredict” to calculate oral cancer risk, guide clinical treatment recommendations, and streamline patient management, thereby improving clinical service delivery. The tool provides enhanced support to clinicians by identifying patients who have a high risk of oral cancer, enabling timely treatment and ongoing close follow-up. Patients who are identified as low-risk can avoid unnecessary surgery and instead follow a biannual monitoring schedule and receive counselling and supportive therapies.
Patients with oral cancer.
The OralCancerPredict tool was designed by Professor Su’s research team in 2021 to estimate cancer risk for patients with unusual white or red precancerous lesions in their mouth, known as oral potentially malignant disorders, such as oral leukoplakia (OL) and oral lichenoid diseases (OLD). The tool examines various data from OL and OLD patients, including age, health issues, lesion appearance, and biopsy findings, to calculate how likely the lesions are to become cancerous over time.
The tool’s underlying AI model was trained on data from a large group of Hong Kong patients with OL and OLD, and validated using data from patients in the United Kingdom and Nigeria. In a 2023 article in the Journal of Big Data, the research team showed that OralCancerPredict had an average accuracy of 94% and was more effective than current risk assessment methods.
“Our tool is freely available on the internet for use by all dentists and doctors everywhere. With a united goal of advancing cancer care, the tool brings precision, confidence, and earlier intervention to clinical decision-making, potentially mitigating oral cancer burden and saving lives,” said team member Professor John Adeoye, Assistant Professor in Digital and Precision Dentistry.
The Faculty AI Clinic welcomes all patients in Hong Kong with OL or OLD. Use of the AI web tool to support risk assessment and treatment decisions does not increase treatment costs, as the AI technology is free and does not require additional tests. However, patients wishing to access this service must first obtain a referral from a general dental or medical practitioner.
From left: Professor Liwu Zheng, Professor John Adeoye, Professor Richard Su, Dr Joanna Ko, Professor Jane Pu