Inspiration
In a hard-working society like ours, we tend not to be aware of essential signs our body sends us, telling us to take a break. There are not many convenient ways to explicitly tell these cumulative signs. We need a tool to remind us. This is why we decided to develop an app for all the people at risk for accumulating these signs. People who need to keep track of the early signs of a detrimental health issue; are our major target audience.
What it does
Our app detects the colouration of the sclera (white part of the eye) and indicates the severity of the eye condition. If the vividness of the colours is detected, the app will suggest the user either take a rest or go see a doctor. It also keeps track of the eye conditions for past recordings, allowing the user to understand the trend of eye conditions according to their daily lives.
How we built it
For the main programing language for the app, we used HTML and CSS. We implemented the YOLOv4 Convolutional Neural Network (CNN) model in Python to first detect the sclera from the image by selecting white regions in the image. From this region, we performed a pixel analysis on RGB values for the red and yellow regions of the sclera to determine the levels of two intensities of red (bloodshot/redness) and yellow.
Challenges we ran into
There is still a machine learning component that has not been solved due to a lack of data samples. Also, there needs to be an improvement in generalizing the lighting when taking pictures as the lighting varies the colour and scrambles eye detection.
Accomplishments that we're proud of
We are proud of involving high levels of coding in a such short period of time.
What we learned
It started off as an idea based on our experience, but we have learned a lot in terms of medical information related to our eyes. Though we knew that eyes are the first indication of sleep deprivation, it was new for us to find out that diseases like diabetic retinopathy or Fuchs’ dystrophy can be diagnosed by the condition of our eyes. In addition, it was interesting to know how liver, gallbladder, and pancreas are all interconnected to each other and a dysfunction of at least one of them is also directly shown in the eyes. Apart from new medical information, we learned how we have been so ignorant of the hints that our bodies have given us, and thus constant check-up of our body is mandatory for all of us. Although the world moves fast and it is just merely impossible to perfect in every aspect of our lives, through this project, it has been a great opportunity for us to remind ourselves again that health is and should always be on top of our priority.
What's next for EYE2EYE: Can EYE See?
The next step for EYE2EYE is to enhance the examination process with more sample data and the implementation of analyzing different colors, chromas, and brightness of eyes. Once the development of the app is enhanced, we can communicate with the doctors to discuss the extent of credibility of the app so that it may be utilized as an easily accessible bridge between patients and doctors.
Log in or sign up for Devpost to join the conversation.