CEO and Founder

The Association for Research in Vision and Ophthalmology (ARVO) Annual Meeting is the most anticipated research conference in the ophthalmology world. Traditionally held every year in late April or early May, the conference was canceled this year due to the COVID-19 pandemic.

We had planned to present three abstracts at the conference and would also have joined the exhibitors. Although the abstracts will be published online in the coming months, we wanted to highlight the findings of the diabetic retinopathy screening study we conducted in Santa Cruz de la Sierra, Bolivia, in collaboration with Retina Global and Volk Optical.

The Santa Cruz Diabetic Retinopathy Artificial Intelligence (SC-DRAI) team. Left to right: Ezequiel Lukin, Simon Barriga, Dr. Christian Diaz, Dr. Olivia Baldivieso Dr. Eva Rosita Dewi, Dr. Jimmy Borda, and Dr. Rajat Agrawal.

In that study, we used EyeStar, our artificial intelligence (AI) software to detect diabetic retinopathy. EyeStar is being used in Mexico at Clínicas del Azúcar, a network of diabetes clinics, where we have screened over 30,000 patients for diabetic eye disease.

Study objectives and results

The objective of this study was to assess the clinical feasibility of diabetic retinopathy screening using the EyeStar AI software paired with a handheld retinal camera, as compared to image evaluation performed by retina specialists, in Bolivian patients. A country like Bolivia, which supports a growing population of patients diagnosed diabetes and with only 2.1 ophthalmologists per 100,000 people, urgently needs cost-effective ways to detect diabetic retinopathy that do not depend on specialists. A low-cost handheld retinal camera paired with AI software could supply preventative eye care to those who are most in need.

We enrolled 1,005 subjects in the study, of which 76 (8.6%) had referable diabetic retinopathy or macular edema. When compared to the adjudicated reads of three retina specialists, EyeStar’s sensitivity reached 92.1% with a specificity of 82.9%. In other words, the EyeStar system can detect nine in ten people with sight-threatening disease while referring only one out of five patients for further evaluation.

Even though we collected images with portable cameras, the EyeStar software was able to process 95% of the cases, showing high resilience to variations in image quality.

Our study shows that EyeStar fulfills the need for diabetic retinopathy screening in countries with few eye-care specialists and increasing rates of diabetic eye disease. Look for the full results of the study in the upcoming Investigative Ophthalmology & Visual Science annual meeting edition, and we look forward to seeing you next year in San Francisco!

CEO and Founder

Would you like to know more about VisionQuest Biomedical?

Join our e-mail list to receive news relatedto Artificial Intelligence and VisionQuest Biomedical. Our company is growing, and we look forward to sharing updates with you!