Access our diverse and vast retinal image database for your research needs.
Our retinal image database is comprised of over 5 million retinal images of diverse populations with various degrees of diabetic retinopathy. Because of the diversity, it helps algorithms recognize the different retinas that exist in real world settings.
Major retinal algorithm development programs worldwide use our database to find new and innovative solutions for diabetic retinopathy.
Contact us to discuss how you can access and use our database.
Kaggle Competition: As computers become more adept at recognizing patterns, the California Health Care Foundation and EyePACS wondered if they could recognize signs of diabetic retinopathy. We reached out to Kaggle to host an algorithm competition in which teams used EyePACS retinal images to train algorithms to note signs of the disease.
Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs: In this research paper, we provided our retinal image database to Google to validate a deep learning algorithm for detecting diabetic retinopathy. Although the algorithm had high sensitivity and specificity, further research is essential to conclude if it can be used in a clinical setting and if it can improve care.
Big Data Takes on Diabetic Retinopathy: EyePACS believes that artificial intelligence can be used to help prevent blindness but it needs guidance to be patient-centric.