AI-based Eye Diagnosis Research Featured in NewsWorker Korea

It’s my pleasure to inform that our AI-based Eye Diagnosis research was featured in “NewsWorker”.

Here is the link: https://n.news.naver.com/article/030/0003036257

Here is the English translation:

“Dry and Stiff Eyes, Could it be Dry Eye Syndrome?… Now Diagnosed with Artificial Intelligence”

The article discusses a breakthrough in diagnosing dry eye syndrome using artificial intelligence to analyze meibomian gland loss. GIST’s Department of Biomedical Engineering team led by Professor Euiheon Chung, in collaboration with Professor Ho Sik Hwang from the Department of Ophthalmology at Catholic University (Yeouido St. Mary’s Hospital), has developed technology for automatic quantitative analysis of meibomian gland images using AI.

The research team has made all materials and analysis data publicly available online, allowing researchers in related fields to attempt new AI technologies and verify their results.

Dry eye syndrome is broadly divided into aqueous-deficient dry eye, where tear production is insufficient, and evaporative dry eye caused by insufficient oil layer production. Excessive tear evaporation is one of the most important causes of dry eye syndrome.

The meibomian glands, located in the eyelids, are responsible for forming the oil layer of the tear film. Evaporative dry eye syndrome is mostly caused by meibomian gland loss or dysfunction.

Using real patient meibomian gland images and their newly developed deep learning model, the research team achieved faster and more accurate diagnostic results than physicians. They used 1,000 images from Yeouido St. Mary’s Hospital for training, with 800 images used for the deep learning model and 200 for validation. When tested on 600 additional images from Korea University Ansan Hospital, the AI continued to demonstrate superior speed and accuracy compared to clinical experts.

The research was led by Professor Euiheon Chung (corresponding author) from GIST and Professor Ho Sik Hwang (corresponding author) from Catholic University Medical School, with support from GIST GRI and the Korea Health Industry Development Institute.

It was presented by: Joon-seong Cho Published on: August 11, 2022, 11:03 AM Source: NewsWorker (뉴스워커)