The RetinaCheck project

The RetinaCheck project is a Sino-Dutch collaborative project, aiming to set up a large-scale screening program for finding early eye damage in diabetes patients (diabetic retinopathy) in the province of Liaoning, Northern China.

Diabetes is reaching epidemic proportions worldwide, especially in Asia due to fast lifestyle changes and genetic factors. It is estimated that currently 11.6 % of the Chinese population has diabetes-2, the most of any country. Diabetic retinopathy is also the main cause of newly formed blindness in the working population, leading to high societal costs. Early detection is the key to prevention and successful treatment. Many cases still go unnoticed and are not treated in time, especially in rural areas.

The technology is based on fully automatic quantitative analysis of high-resolution images of the retina, made with a fundus camera. The software is based on brain-inspired geometric methods, and convolutional neural networks (deep learning).

The departments of Biomedical Engineering at Eindhoven University of Technology (TU/e, Eindhoven, the Netherlands) and Northeastern University (NEU, Shenyang, China) supply the computer-aided diagnosis (CAD) software. The validation of the algorithms is done with clinical data from Shengjing Hospital, Shenyang, the Maastricht Study, Netherlands, and public databases. The acquisition of field images is done through He Eye Care, Shenyang, and collaboration hospitals. The cloud computing and storage services are serviced by Neusoft Inc.

An example of diabetic retinopathy (DR). Many micro-bleeds, vascular changes, and in advanced (proliferative) stages also angiogenesis. The CAD software quantifies the biomarkers, and aims for early detection of DR, even before clinical symptoms appear.

International collaboration

• The Departments of Biomedical Engineering of Eindhoven University of Technology (TU/e) in the Netherlands, and of Northeastern University (NEU) in Shenyang, China, have already for many years a successful collaboration with a joint Sino-Dutch Biomedical and Information Engineering School (BMIE) in Shenyang, with many exchanges of staff and students. Shenyang is home of Neusoft, China’s largest medical imaging and software industry, and the Philips-Neusoft research lab. The BMIE School of NEU is now the largest in Liaoning, with over 800 BME students.
• He Shi Eye Care has taken the initiative for the screening program. It is one of the largest eye care providers in Shenyang / Liaoning, covering 11 eye hospitals and He University, a private university focusing on Ophthalmology. 200 Eye Care centers are projected for the province of Liaoning to participate.
• Maastricht University Eye Clinic is famous for the large ‘Maastricht Study’, an extensive phenotyping study including 5000 healthy persons and 5000 diabetes patients. A wide range of parameters (blood, fundus, MRI, lifestyle, etc.) is cross-correlated for causal relations.
• The high-tech company EasyScan BV in The Hague, the Netherlands, is developer of innovative and low-cost laser-scanning fundus camera’s.

Beijing 24-03-2014: In the presence of minister Jet Bussemaker and dr. Hans de Groene (NWO) the Memorandum of Understanding is signed by prof. Wei He and prof. Bart ter Haar Romeny.

Partners collaborate in the Sino-Euro Vision and Brain Institute.

Beijing 09-09-2013: Prof. ter Haar Romeny, NL ambassador mr. Aart Jacobi, minister Edith Schippers (VWS) and i-Optics’ CEO dr. Jeroen Cammeraat.

Funding

The project is funded by:

•  The 合 Hé Programme of Innovation Cooperation (PoIC), a Sino-Dutch partnership in research and innovation by Netherlands Organisation for Scientific Research NWO (1 PhD student, 1 Postdoc, travel, 2 fundus cameras, software support)  and the Chinese Ministry of Science and Technology MoST (equipment, infrastructure, manpower)
•  The European Foundation for the Study of Diabetes, Lilly Programme (2 postdocs, 2 fundus cameras),
•  Northeastern University (staff, students)
•  TU/e (staff, students)
•  The Chinese Scholarship Council (PhD student 4y),
•  MAnET – EU Marie Curie Initial Training Network (PhD student 4y, postdoc 3y)
•  i-Optics BV (PhD student 4y),
•  He Shi Eye Care (full screening infrastructure),
•  Maastricht University Eye Clinic (Maastricht Study),

Innovative Technology

The design of effective algorithms is inspired by recent findings of how our visual system works. Modern optical and advanced MRI brain research gives more and more insight in the mathematical functions in the brain. These advanced contextual and brain-inspired algorithms for CAD are developed in close collaboration with dr. Remco Duits at TU/e, who recently was awarded an European Research Council (ERC) grant of 1.3 M€ for this innovative technology, and several other European partners in a recently awarded EU FP7 People-ITN project (MANET).

Validation Study

A crucial step is the validation of the developed image analysis algorithms for their classification power. We are setting up a large validation study with patients of the He Shi Eye Hospitals, of China Medical Universiy’s Shengjing Hospital and or normal subjects measured at the City Health Check Center in Shenyang. Many other details of the patient are registered, so they can be correlated with the disease progress and the features found.

We have much benefit from the Maastricht Study (Maastricht, the Netherlands), world’s largest phenotyping study to find relations between the progress of diabetes and many measured parameters.

One-Stop-Shop

The aim is to develop a one-stop-shop for at least three common eye diseases: diabetic retinopathy, glaucoma and age-related macular degeneration (AMD).

Next steps in this long-term project will be the inclusion of 3-dimensional (OCT) and very high resolution scanning techniques (adaptive optics) of the retina, and the establishment of a dedicated 3T MRI facility for the clinical and research study of the visual system in the brain. This will be done in close collaboration with prof. Mark Haacke, Detroit, USA. TU/e will collaborate in making available software and analysis for interactive visualization of the connectivity (nerve fiber bundles) of the visual system in the brain (see for some examples here). The focus on the visual system is unique.

Impact

The combination of a large and experienced clinical partner with two high ranking engineering universities and associated high-tech partners, is an excellent formula for attacking this major epidemic diabetes problem in China, and its associated prevention of blindness. Vision for Vision, Technology for Healthcare.