Abstract
Purpose::
To evaluate the accuracy of a hand-held infant keratometer (IK4), as compared to keratometry measurements obtained with the Nikon Retinomax K+ autorefractor in preschool children who are members of a Native American Tribe with a high prevalence of astigmatism.
Methods::
Subjects were 19 preschool children who were enrolled in Head Start on the Tohono O’odham Reservation during the 2006/07 school year. Keratoscope video images for each subject were obtained using a single-ring (200 mm diameter) keratoscope with 12 infrared LEDs attached to a telecentric telephoto lens designed for use with a Sony TRV460 Digital-8 camcorder. Each subject also had corneal astigmatism measured by the Retinomax K+ autorefractor. Custom image analysis software was written to score IK4 image data. The software automatically scans captured video, analyzing each frame for keratometry, pupil size, alignment and image motion. The software then selects the best nine images to provide mean estimates of corneal curvature and corneal astigmatism. These results are then compared to the Retinomax estimate for each subject.
Results::
Scoreable IK images were obtained for 18 of 19 children and Retinomax measurements were obtained for all 19 children. Retinomax keratometry indicated that corneal astigmatism ranged from 0.50 to 5.13 D. 16 children (89%) had corneal refractive astigmatism ≥ 1.00 D and 12 (67%) had corneal refractive astigmatism ≥ 2.00 D. The mean difference between measures of corneal astigmatism with the IK4 and Retinomax was significant (0.37 D, SD=0.56 (p <0.02)), with the IK showing slightly higher values. The two measures of astigmatism differed by < 0.50 D for 72% of subjects.
Conclusions::
The IK provided slightly higher estimates of corneal astigmatism than the Retinomax in this highly astigmatic sample of children.
Keywords: refraction • astigmatism • imaging/image analysis: clinical