Real-time, computer-assisted quantification of plus disease in retinopathy of prematurity at the bedside

Michelle T. Cabrera, Sharon F. Freedman, Mary Elizabeth Hartnett, Sandra S. Stinnett, Bei Bei Chen, David K. Wallace

Research output: Contribution to journalArticle

6 Scopus citations

Abstract

BACKGROUND AND OBJECTIVE: Plus disease is the primary indication for retinopathy of prematurity (ROP) treatment, but in borderline cases ophthalmologists may struggle to judge whether it is present. ROPtool is a semi-automated computer program that objectively assesses plus disease by measuring retinal vascular tortuosity and width. This study determined ROPtool's bedside diagnostic accuracy concurrent with ROP screening. PATIENTS AND METHODS: ROP screening examinations were recorded using Keeler video indirect ophthalmoscopy. A masked operator traced images in ROPtool at the bedside, comparing ROPtool's plus diagnosis to the examiner's clinical judgment. RESULTS: Four hundred sixty-four examinations (129 eyes of 65 infants) were performed. ROPtool's sensitivity, specificity, and area under the receiver operating characteristic curve for plus diagnosis was 71% (CI: 38%-100%), 93% (CI: 89%-98%) and 0.87, and for pre-plus or worse was 68% (CI: 51%-85%), 82% (CI: 77%-86%) and 0.81, respectively. CONCLUSION: ROPtool can provide a real-time second opinion of plus disease at the bedside. Image enhancement technologies may further improve ROPtool's diagnostic accuracy.

Original languageEnglish (US)
Pages (from-to)542-548
Number of pages7
JournalOphthalmic Surgery Lasers and Imaging Retina
Volume45
Issue number6
DOIs
StatePublished - Nov 1 2014
Externally publishedYes

ASJC Scopus subject areas

  • Surgery
  • Ophthalmology

Fingerprint Dive into the research topics of 'Real-time, computer-assisted quantification of plus disease in retinopathy of prematurity at the bedside'. Together they form a unique fingerprint.

  • Cite this