Purchase this article with an account.
C. M. Wilson, K. Wong, K. D. Cocker, J. Ng, A. L. Ells, A. R. Fielder; Could Computerized Analysis of Retinal Vessel Tortuosity Predict Retinopathy of Prematurity (ROP) Requiring Treatment?. Invest. Ophthalmol. Vis. Sci. 2009;50(13):5723.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Retinal vessel tortuosity has been recognised as a sign of advancing retinopathy of prematurity (ROP) since the first description of retrolental fibroplasia over 50 years ago. Recently, digital image analysis techniques have enabled quantification of this pathological change. This study quantifies tortuosity comparatively in infants requiring and not requiring treatment for ROP.
130 degree RetCam II (Clarity Medical Systems Inc., Pleasanton, CA) images from preterm infants undergoing routine screening for ROP were selected. Infants scheduled for subsequent laser treatment for ROP were noted. Tortuosity values of one artery and one vein in each quadrant of the retina were measured using Computer Assisted Image Analysis of the Retina (CAIAR - Imperial College London). For analysis, a Mann- Whitney U test of tortuosity of both vein and artery metrics combined was carried out. Tortuosity of veins and arteries separately was also analysed.
A total of 514 vessels from 68 images were analysed. 22 infants received laser treatment for ROP subsequent to imaging, 46 infants were in the non-treatment group. Mean tortuosity values for veins and arteries combined in treated infants was 9.07 (SD ±3.84) and in the non-treated infants was 5.98 (SD ±3.17). There was a statistically significant difference in the vessel tortuosities in infants who received treatment, compared to those who did not (p=0.001). Further, there was a statistically significant difference between treatment and non-treatment groups when veins and arteries were analysed separately (veins p=0.017, arteries p=0.004). Tortuosity was consistently higher in arteries (7.04 ±6.64 SD non-treatment, 11.13 ±6.18 SD for treatment) than in veins (4.94 ±2.23 SD without treatment, 7.06 ±3.89 SD for treatment).
Advances in digital image analysis can provide accurate, repeatable measures of arterial and venous tortuosity. In the future it may be possible to define a cut-off value for vessel tortuosity indicative of infants at risk of developing treatable ROP.
This PDF is available to Subscribers Only