AI-Derived Vascular Severity Score May Standardize ROP Treatment

Premature Baby Retinopathy
Retinopathy In Premature Infant. Retinal Fluorescein Angiogram Showing Hypoxia Of The Retinal Periphery. Recolorized, Black And White Image. (Photo By BSIP/UIG Via Getty Images)
Researchers developed an artificial-intelligence derived vascular severity score to aid in the diagnosis of retinopathy of prematurity.

This article is part of Ophthalmology Advisor’s Focus on Retina coverage from the 2021 meeting of the American Academy of Ophthalmology, held in New Orleans from November 12 to 15, 2021. The team at Ophthalmology Advisor will be reporting on a variety of the research presented by the retinal experts at the AAO. Check back for more from the AAO 2021 Meeting.


An artificial-intelligence (AI)-derived vascular severity score may help standardize treatment of patients with retinopathy of prematurity (ROP), according to research presented at the American Academy of Ophthalmology 2021 annual meeting, held in New Orleans from November 12-15.

Investigators included 267 eyes from the imaging & informatics in ROP (i-ROP) cohort study. Each eye had an independent reference standard diagnosis (RSD). The team utilized the i-ROP deep learning algorithm to provide quantitative vascular severity scores for eyes that have been diagnosed with and without type I ROP. 

The researchers found that in 11% of cases physicians documented treatment for cases less than type I ROP. When an RSD was assigned, researchers noted that 52% of eyes were treated for less severe disease. Furthermore, type I ROP eyes were found to have a higher vascular severity score of 8.3±1.2 in comparison to those with less severe disease which had a score of to 5.7±2.3 (P <.005).

“These findings have implications for clinical diagnosis, application of clinical trial data and the need for objective metrics of disease in ROP,” according to the presenters.

Visit Ophthalmology Advisor’s conference section for the complete Focus on Retina coverage from the AAO 2021.



Nguyen T-T, Kalpathy-Cramer J, Chiang MF. Evaluation of AI for standardization of treatment in ROP. Paper presented at: The American Academy of Ophthalmology 2021 Annual Meeting; November 12-15, 2021; New Orleans. Abstract PA013.