A team of Melbourne-based IBM researchers have trained the company’s Watson AI to detect abnormalities in photographs of eyes, specifically retinas, in a development that could one day speed identification of patients at risk of eye diseases like glaucoma.
The research, which began in 2015, has focused on automating a lot of the manual processes doctors conduct to identify signs of potential retinal abnormalities and has achieved several milestones so far.
The technology can distinguish between left and right eye images, evaluate the quality of retina scans and identify possible indicators of glaucoma.
Image analytics technology was used to scan 88,000 retina images and identify, through deep learning techniques, anomalies in the eye. The results showed Watson was able to measure the optic cup-to-disc ratio, and determine key glaucoma signs with a statistical performance that was up to 95 percent accurate.
Melbourne University’s Dr Peter van Wijngaarden, principal investigator at the Centre for Eye Research Australia said at least 150,000 Australians were estimated to have undiagnosed glaucoma, with that number expected to rise in line with the ageing population.
“It is critical that every Australian has access to regular eye examinations throughout their life so that diseases like glaucoma and diabetic retinopathy can be detected and treated as early as possible,” he said.
“There is a real need for resources that allow all Australians to access regular eye examinations and the development of image analytics and deep learning technology will provide great promise in this area.”
As the research continues, it is expected to investigate the diagnosis of other eye diseases, such as such as diabetic retinopathy and age-related macular degeneration.
Dr Joanna Batstone, IBM Research Australia’s vice president and lab director, said this kind of medical image analysis had the capacity to change the delivery of healthcare services.
“Medical images represent a rich source of data for clinicians to make early diagnosis and treatment of disease, from assessing the risk of melanomas to identifying eye diseases through the analysis of retinas. Cognitive technology holds immense promise for confirming the accuracy, reproducibility and efficiency of clinicians’ analyses during the diagnostic workflow,” she said.
Globally, IBM Research is also combining data analysis with medical images to investigate identification of diseases such as melanoma and breast and lung cancer.