About

In 2019, it was estimated that there were 2.2 billion people experiencing vision impairment and blindness. According to the Rapid Assessment of Avoidable Blindness (RAAB), the prevalence of blindness in Indonesia reached 3%, making Indonesia the country with the highest rate of blindness in Southeast Asia. The global incidence of keratitis is approximately 0.4 to 5.2 per 10,000 people annually. The incidence of keratitis is higher in developing countries compared to developed countries. The presence of general practitioners in primary care facilities can be crucial in detecting cases of keratitis. However, the limited availability of diagnostic support tools often results in keratitis patients not being treated properly and sometimes being referred late.

Efforts to combat cases of keratitis require an approach that empowers doctors through increasing knowledge and awareness by developing a simple keratitis detection tool that can be used independently by primary care physicians. This tool should not require expensive equipment and should be easy to use to aid in early detection and diagnosis of keratitis cases. Up to now, keratitis detection tools using image classification from smartphone cameras have not been widely implemented. Previous research studies have limited amounts of training and testing data. Therefore, there is a need to develop a keratitis detection model from smartphone photos using Convolutional Neural Network (CNN) methods.

This research is expected to produce a set of keratitis detection tools from smartphone camera photos integrated into a smartphone application, which can be used as an early detection tool for keratitis in primary care settings. The development of a keratitis diagnostic tool based on Artificial Intelligence (AI) technology can generate output in the form of mapping the distribution of keratitis cases in the population. This can enable preventive interventions through referral systems and definitive actions such as surgery to prevent blindness due to permanent corneal damage and infection.