Skip to content

Deep Learning for Medical Diagnosis

Deep learning used to diagnose diabetic retinopathy
Researchers increased image processing speed 57x using Julia.

Diabetic retinopathy is an eye disease that affects more than 126 million diabetics and accounts for more than 5% of blindness cases worldwide. Timely screening and diagnosis can help prevent vision loss for millions of diabetics worldwide, but many of them lack access to health care.

IBM and Julia Computing analyzed eye fundus images provided by Drishti Eye Hospitals, which provides eye diagnosis and care to thousands of rural Indians.

Drishti founder and CEO Kiran Anandampillai explains:

India is home to 62 million diabetics, many of whom live in rural areas with limited access to health facilities. Timely screening for changes in the retina can help get them to treatment and prevent vision loss. Julia Computing's work using deep learning makes retinal screening an activity that can be performed by a trained technician using a low cost fundus camera.

By combining Julia’s superior speed and performance with IBM’s Power8 server and NVIDIA Tesla K80 GPU accelerators, researchers increased image processing speed 57x – a dramatic improvement.

According to Julia Computing CEO Viral Shah,

IBM Power provides 2-3x more memory bandwidth combined with tight GPU accelerator integration to create a high performance environment for deep learning with Julia.

The results were presented by IBM and Julia Computing at SC16, the world’s most important supercomputing conference and more details are available on the Julia Computing blog.