Today many Californians with diabetes do not get screened for diabetic retinopathy, a sight-threatening complication of the disease, due to the related costs and limited access.
In the future this might change as researchers develop computer algorithms that could screen a retinal image for diabetic retinopathy as well as human clinicians, thereby making readings faster, more cost-effective, and potentially more accurate.
To catalyze advancement in this field, CHCF partnered with Kaggle, a competition platform for predictive modeling and analytics.
The Diabetic Retinopathy Detection competition drew on the expertise of computer scientists, statisticians, engineers, and data miners from all over the world. The top three teams used image classification, pattern recognition, and machine learning to develop automated diabetic retinopathy detection models that are each on par with human performance.