Big Data Competition

Supported by the Department of Medical Services of the Ministry of Public Health of Thailand, the 4th APTOS Big Data Competition focuses on the generation of optical coherence tomography (OCT) images from color fundus photography. Paired datasets of fundus photos and OCT images of over 2,000 patients with different diabetic maculopathies and glaucoma will be used.
 

Description

Despite the many benefits of the fundus photo, the fundus imaging system sees what the human eye can see. With this technology, an ophthalmologist or optometrist detects only pathologies that are visible to human eyes. Many diseases at the early stages are almost invisible to even an experienced optometrist or ophthalmologist. Most retinal abnormalities progress with age and develop slowly and gradually, so diagnosing them is pretty difficult.
 
On the other hand, modern OCT examination allows ophthalmologists to get images with a reasonably high resolution. Modern OCT image interpretation, with the 3D volumetric data and detail of the retinal layers that it captures, also allows physicians to detect the warning signs of the disease, classify hundreds of pathologies, and re-monitor images to track the progression of pathologies.
 
We have heard so much about what generative AI is capable of doing. What if it can help generate more detailed, yet expensive, OCT images from affordable fundus photos? In this synchronous competition, you’ll build a machine learning model to generate OCT images from color fundus photos. You’ll work with fundus photos of over 2,000 patients collected in hospital settings. If successful, you will make quality eye care and diagnosis accessible to patients in low-and middle-income countries. Such AI tools could enhance teleophthalmology and remote screening, providing valuable structural information without the need for specialized OCT equipment, which can certainly benefit patients with limited resources.

Bitnami