New York, Feb 23 : A novel image-based diagnostic tool, developed using artificial intelligence (AI) and machine learning techniques, may potentially speed up diagnoses and treatment of patients with retinal diseases and pneumonia among children, researcher say.
The findings showed that the new tool uses big data and AI to not only recognise two of the most common retinal diseases -- macular degeneration and diabetic macular edema -- but also to rate their severity.
It can also distinguish between bacterial and viral pneumonia in children based on chest x-ray images.
"Macular degeneration and diabetic macular edema are the two most common causes of irreversible blindness but are both very treatable if they are caught early," said Kang Zhang, Professor at the University of California-San Diego.
"Deciding how and when to treat patients has historically been handled by a small community of specialists who require years of training and are concentrated mostly in urban areas."
"In contrast, our AI tool can be used anywhere in the world, especially in the rural areas.
This is important in places like India, China and Africa, where there are relatively fewer medical resources," Zhang said.
For the study, published in the journal Cell, the team studied over 200,000 optical coherence tomography (OCT) images using a technique called transfer learning, where knowledge gained in solving one problem is stored by a computer and applied to different but related problems.
The researchers also used occlusion testing, which allowed them to show areas of greatest importance when reviewing the scan images.
"Machine learning is often like a black box where we don't know exactly what is happening," Zhang said.
The researchers then compared the diagnoses from the computer with those from ophthalmologists who reviewed the scans.
The results showed that the tool "could generate a decision on whether or not the patient should be referred for treatment within 30 seconds and with more than 95 per cent accuracy", Zhang said.
Besides eye diseases, the tool was able to differentiate between viral and bacterial childhood pneumonia with greater than 90 per cent accuracy.
It can also discern between cancerous and non-cancerous lesions detected on scans, Zhang said.