PolyU and OPTICAL 88 launch joint research project to promote smart clinics recruiting high-risk diabetic patients for new AI-enabled self-testing retinal fundus camera

It is estimated that in Hong Kong approximately 10% of individuals aged 35 to 64 suffer from diabetes, with 30% of them developing diabetic retinopathy (DR) and facing risk of blindness. This results in substantial demand for annual retinal examinations, creating a significant healthcare challenge. The current shortage of manpower in the public healthcare system and long waiting times are delaying timely and appropriate treatment. The “Smart primary healthcare and eyecare service: From data to algorithms and real-world solutions” project, funded by the Hong Kong Government’s Global STEM Professorship Scheme, has supported The Hong Kong Polytechnic University (PolyU) and OPTICAL 88 in a DR screening project for diabetic patients which aims to relieve the pressure on the public healthcare services’ primary eye care sector. Working on the project since March 2024, Prof. Mingguang HE, PolyU Chair Professor of Experimental Ophthalmology of the School of Optometry, Global STEM Scholar under the Global STEM Professorship Scheme and Henry G. Leong Professor in Elderly Vision Health, has invented a low-cost, portable, self-testing retinal fundus camera with a custom artificial intelligence (AI) system to conduct DR screening for eligible personnel.

Portable, self-testing retinal fundus camera proves high accuracy and efficiency

The camera with custom AI and mechanical automation systems developed by Prof. He and his team have not only resolved the limitations of traditional retinal fundus camera equipment, but has also enabled small-scale and low-cost self-testing medical screening, revolutionising the primary eye care sector. Retinal and optic nerve diseases, such as diabetic retinopathy (commonly known as diabetic eye), age-related macular degeneration, glaucoma, retinopathy of prematurity and others all have ocular manifestations in the retinal fundus. Retinal fundus photography can, therefore, assist in diagnosis.

The team’s portable, self-testing retinal fundus camera with a custom AI system has a sizeable database of clinical and fundus image data that is connected to a cloud-based analysis system. Leveraging this extensive database, the team has developed a deep learning model that can automatically identify and analyse retinal fundus images, with a proven accuracy of 98.9% in detecting referrable DR. The application of this AI fundus photography technology in the primary eye care sector is expected to help resolve issues of high healthcare cost and shortage of manpower.

Prof. He said, “The self-testing retinal fundus camera with custom AI system will not only improve the efficiency and accuracy of primary eye care, but will also promote self-testing amongst the Hong Kong public, with the hope that more organisations will be able to develop smart clinics, introducing such equipment at affordable prices and helping to relieve the burden on the public healthcare system.” Prof. He adds that the smart clinic research conducted in collaboration with OPTICAL 88 will facilitate the use of AI-backed, low-cost portable retinal fundus cameras for DR screening in the primary care system, improve accessibility to healthcare amongst the Hong Kong public and ensure more patients receive the timely diagnosis and treatment of ocular diseases.

Recruitment for the evaluation of the benefits of self-testing

OPTICAL 88 is the first optical chain to employ the self-testing retinal fundus camera with custom AI and mechanical automation, with the systems installed in two of its retail shops. It is recruiting people aged 50 or above and diabetic patients aged 18 or above who have not had an eye examination in the past 12 months to participate in the AI Diabetic Eye Screening Research Project. Participants will be assessed and then undergo AI diabetic eye screening, alongside a consultation with OPTICAL 88’s optometrists. They are required to complete a questionnaire survey and conduct a telephone interview three months later. The study will evaluate the self-testing model; test the diagnostic accuracy of the system; determine screening rates and patient compliance; as well as assess the cost-effectiveness of the self-testing model in comparison with traditional methods by studying the perceptions, confidence, experience and satisfaction of primary healthcare providers (including general practitioners and optometrists), support staff and clinic managers.

Mr Joshua S. Kanjanapas WONG, Director of OPTICAL 88, said, “The vision health of the Hong Kong public has always been a priority for OPTICAL 88. We have invested very considerable resources in educating the public on the importance of regular and comprehensive eye examinations. Many ocular diseases do not show symptoms at the early stages, meaning that the prime window for treatment may be missed by the time symptoms become apparent. OPTICAL 88 keeps abreast of the latest technological developments and frequently introduces appropriate healthcare products to offer professional and convenient services to the public, aiming to promote the advancement of optometry in Hong Kong.”

“We are delighted to join hands with PolyU in the AI Diabetic Eye Screening Research Project. The Project’s objectives align perfectly with our company’s ethos, which is to offer efficient assessments of high-risk individuals for early detection and timely referrals so they can seek the professional help and treatment that they need. Meanwhile, the data collected from the Project will evaluate the benefits of AI technology in the primary eye care sector and facilitate the development of smart clinics, so as to foster the future development of ophthalmology,” Mr Wong added.

For enquiry and application of the recruiting scheme, please email to aiclinic@polyu.edu.hk.

***END***