UNAIR NEWS – Now, cataract eye detection is as easy as touching the android phone screen. A cataract examination application based on machine learning has just been invented. This application is able to analyze and detect cataract eye in seconds with an android phone.

A startup company www.cekmata.com was crowned as the champion in the competition held by Telkomsel, The Next Dev 2017 for E-Health category. The application is the result of collaboration of UNAIR Faculty of Medicine student Caesar L. Givanni with two friends of his, Ivan Sunarso and Sylvester Albert Samadhi.

Caesar who is currently undergoing Specialist Doctor Education Program (PPDS) for Internal Medicine RSUD Dr. Soetomo-FK UNAIR claimed that this application’s jargon Selfie-Check Healthy will simplify screening / examination for optic health, especially for early detection of cataract.

Patients with cataract will recover faster if the symptoms can be noticed early on. Unfortunately, access to eye screening is still low, while awareness for eye screening is also limited in society. As a result, new cataract are known after acute conditions.

“Nearly 4 million people in Indonesia are exposed to cataract. WHO estimated that by 2020, every one minute Indonesian will be blind because of cataract. The urgency is high, considering cataract is the biggest cause of blindness in the world, “Caesar said.

Because of the problem, the idea to make an application of cataract examination via mobile phone came up. These three comrades combined three disciplines that each of them master, Medicine, IT (machine learning), and marketing.

Chief Technology Officer cekmata.com Albert combined the technology machine learning with cataract screening concept. Albert illustrated that in the making of the application, the computer was taught how to distinguish eyes with cataract and normal eyes, through comparison of data. The more comparative images were input, the more intelligent the computer, the better it is able to analyze and to produce higher analytical accuracy.

“This is the intelligence of computers, the characteristic is quite simple in terms of machine learning. It fits, ” he said.

Abroad, machine learning-based technology is being discussed. This is a new method of analysis that can be applied for various purposes. In addition to be useful in the field of medicine, machine learning method can also be used for online store analysis, stock, business and financial stock predictions.

“Research utilization of learning machine to detect skin cancer is being done abroad. It’s still very new, and Indonesia should not miss it, “said Albert.

Caesar, Ivan, and Albert felt lucky to come up with the idea of ​​making a cataract check application using machine learning system. Because abroad, not many have applied this system into a product.

How to use www.cekmata.com app is fairly easy. Simply by accessing www.cekmata.com, users can immediately take advantage of the eye examination feature in the application. For the screening process, users can direct one pupil directly to the mobile camera , then take a snapshot. This application will detect the presence of cataract in the eyes of users.

This app will directly measure the convidence percentage of the eye images sent by the user. Within seconds, the user can immediately know the results whether his eyes are healthy or not.

When the results of the analysis shows positive result for cataract, then the application will automatically direct the user to perform further examination to the optical clinic or the nearest hospital. To facilitate the user, this application is also equipped with distance information and reputation of the clinic or hospital recommended in the application.

“Just tell the doctor, ‘Doc according to cekmata.com i have cataract‘. The doctor will examine further, “said Caesar.

To support the examination process, Albert suggested that the shots are taken in good lighting. In addition to developing online–based applications , this application can also be accessed offline. So this application can still be accessed even without an internet connection.

“Now anyone can get into early cataract detection. They do not have to ask the parents to take them to the doctor, anyone can access this application, so it can directly detect cataract or not, “said Albert.

The three SMAK 1 St Louis Surabaya graduates claimed the accuracy level of the application can give correct analysis result up to 75-80 percent.

“We still need a lot of data to make the computer cleverer. As more data learned our computer, the more accurate the result is, close to 95 percent, “Albert said.

Meanwhile, Chief Marketing Officer www.cekmata.com, Ivan Sunarso, revealed that the concept of social technopreneurs he experienced in the last two months required supports of many parties. For that, Ivan planned to embrace various cataract care communities, to disseminate the application so it can be accessible by people from urban to rural areas.

They are constantly developing the features of www.cekmata.com by enriching comparable data to be learned by the computer system.

“There is a possibility there will be collaboration with other parties to improve the features of this application. They are currently establishing cooperative access with a number of hospitals and optical clinics, ” he said.

They hoped that the presence of this application can reduce the number of blindness in Indonesia.

“We struggle for this application remain free and accessible to anyone, anywhere. Early cataract eye screening will not be a problem for anyone, “said Caesar.

Author: Sefya Hayu Istighfaricha

Editor: Binti Q. Masruroh