HEIDELBERG RETINA TOMOGRAPH - REPRODUCIBI LITY AND RELATIVE ERROR IN THE MODEL EYE AS A FUNCTION OF THE WIDTH OF THE PUPIL

Background We measured the reproducibility and relative error of the H eidelberg Retina Tomograph (HRT) in a model eye with various widths of the pupil. Methods We put diaphragms of various diameters in front of the lens of a model eye. The volume of an artificial papilla which wa s inserted into the model eye was measured three times. The coefficien t of variation (standard deviation/mean value) was taken as a measure of the reproducibility. The relative error was defined as the absolute value of the quotient of the difference between measured and real val ue. Results The relative error for diaphragms of 1.5 to 5 mm was 7% to 18.4%, for a diaphragm of 1 mm 19.8%, and for a diaphragm of 11 mm 21 .3%. The coefficient of variation ranged from 0.49% to 1.95% for diaph ragms of 1.5 to 11 mm diameter, and was 8.1 for a diaphragm of 1 mm di ameter. Conclusion The HRT is capable of measuring volumes with a reas onable accuracy. The accuracy does not depend on the width of the pupi l over a wide range of pupil diameters. In cases of very small or very much dilated pupils the accuracy and reproducibility declines.