A student in the US has created a 3D-printed lens that can be used alongside a smartphone app that can spot signs of eye disease in photos.
According to IEEE Spectrum, Kavya Kopparapu, a 16-year-old student from Virginia, was inspired to build the device after her grandfather, who lives in India, began showing symptoms of diabetic retinopathy, an eye disease linked to diabetes that can cause blindness.
The condition is preventable, but many of the poorest sufferers will not get diagnosed, and those with more severe forms can go blind.
Kopparapu decided to use her computer science skills to create a system that can cut down a traditionally two-hour exam to a system process done with a smartphone, which uses AI to offer a preliminary diagnosis.
“The lack of diagnosis is the biggest challenge,” she told IEEE Spectrum.
“In India, there are programs that send doctors into villages and slums, but there are a lot of patients and only so many ophthalmologists.”
Called the Eyeagnosis, her device uses AI to identify symptoms in photos, and was put together after consulting ophthalmologists, neuroscientists and machine learning experts.
The team that built the final system includes her brother and a high school classmate, and they used a neural network – a computer network modelled on the human brain – previously built by Microsoft and was able to analyse images.
More than 34,000 retinal scan images from the National Institute of Health’s database in the US were then used to help train the system to spot symptoms.
The final system is also much simpler than traditional testing, which requires a large machine and take up to two hours to complete.
The resulting prototype has already been tested at the Aditya Jyot Eye Hospital in Mumbai on five patients, according to IEEE Spectrum.
Any plans for expansion are yet to be announced, but Kopparapu did present the Eyeagnosis system at the O’Reilly Artificial Intelligence conference in New York in June.
Computer science student Kopparapu has also founded her own coding workshop group – the Girls Computing League.
