Using AI to See What Eye Doctors Can’t

Print Friendly, PDF & Email

This white paper, “Using AI to See What Eye Doctors Can’t,” explains how Voxeleron, a leader in delivering advanced ophthalmic image analysis and machine learning solutions, is extending ophthalmology’s diagnostic horizons with image analysis based on artificial intelligence (AI) models, trained using Dell Precision workstations with NVIDIA GPUs.

In order to process 3D data sets in a deep-learning, convolutional neural network (CNN) model, Voxeleron needed dramatically more computer processing power. Its goal? A tool to help ophthalmologists predict when  patients with dry age-related macular degeneration (AMD) might advance, leading to blindness.

As the world’s population ages, experts expect vision loss and complete  blindness in elderly populations to grow, especially from incurable age-related macular degeneration (AMD). AMD destroys the macula, the area  of the retina responsible for the sharp, central vision necessary to see objects clearly.

Today, AMD affects nearly 200 million people globally, making it the leading cause of blindness in people 60  years and older. The disease has two types: an early dry stage and an advanced wet stage when vision loss and blindness can rapidly follow.

Currently, ophthalmologists use optical coherence tomography (OCT) to generate 3D retinal images to  diagnose and monitor the disease. Unfortunately, they can’t tell which patients will convert from a dry to a  wet stage, so they must monitor them regularly for changes in the condition.

Download this white paper, “Using AI to See What Eye Doctors Can’t” to read more about how Voxeleron saved as much as three months by running their AI models on the Dell Precision 7920 Tower workstation  running NVIDIA Quadro GV100 GPUs versus their original setup.