Retinal diseases are wide-ranging, in their symptoms as well as in severity. Left untreated, some diseases can cause severe vision loss, impairment or even blindness. The importance of proactive treatment and prevention is of vital importance. However, everybody has not that equal opportunity.
In developing countries in particular, a critical issue is a lack of public access and means to specialised medical care. Even if the equipment is there, the expertise required to interpret test results and get the correct diagnosis and prognosis may be absent.
For retinal diseases this is an acute issue. Accurate diagnosis and interpretation of test results are extremely difficulty, especially for inexperienced doctors.
PixelPlex, a development and consulting company focused on blockchain, artificial intelligence (AI) and Internet of Things (IoT) technologies among others, concluded that, through its contacts with medical facilities, it may be able to assist. “We realised that we could actually use AI for the automatic diagnostical aids,” explains Alexei Dulub, founder and CEO of PixelPlex.
AIRA, an artificial intelligence retina analyser, is the result of the attempt. Being a pattern recognition use case, AI was the obvious solution. Although a neural network will ultimately not diagnose better than a medical professional, the undeniable advantage of AI is in analysing large data sets, and evaluating screening results in specific numbers. AIRA can assist a doctor in getting more accurate data for further work.
PixelPlex was originally given a ‘large array’ of data with images from fundus cameras, which contained various symptoms and anatomical structures of the human eye, such as exudates, hemorrhages, and degenerative retinal changes, vessels and optic nerve pathologies. The company said that its own data sets from later research, are to be used in neural network training.
“We created our own data set, which was quite a challenge, and with the help of qualified medical professionals, created the UI data set that trained the AI to detect various diseases and other defects, just based on photographs taken by a fundus camera,” said Dolgov.
The model architecture was based on differences of U-Net, a convolutional neural network developed for biomedical image segmentation. These differences included LinkNet, a light deep neural network architecture for semantic segmentation, and Dilated U-Net, which has been used in other initiatives for assessing the risk of cancer in certain organs.
The photos taken by the fundus camera are sent and analysed by the software created by PixelPlex.The trained neural network is capable of identifying information to decide on diagnoses, then provide medical staff with that information.
“The source of these images, having them analysed, providing the medical diagnosis based on the photographs, and then creating this data set – that was the biggest challenge,” said Dolgov.
Currently, the solution has approximately 85% accuracy. But, the project is still in active development, and it is expected to increase the rate of accuracy to, at least, 95%. PixelPlex observes that AIRA will be able to spot symptoms of disease with such high precision which a regular physician could not reach, and also create mathematical models later to be used to enhance the neural network analysis process.
