Purchase this article with an account.
E. Bentley, K.A. Diehl, P.E. Miller; Measurement of High Resolution Ultrasound Images: Intra-Observer and Inter-Observer Reliablity and Image Reproducibility . Invest. Ophthalmol. Vis. Sci. 2003;44(13):3608.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: The purpose of this study was to evaluate intra-observer and inter-observer reproducibility of measurements and reproducibility of anterior segment images in dogs obtained by high resolution ultrasound. Methods: Intra-and inter-observer reproducibility: Images were obtained from eight eyes of four dogs using a 20MHz transducer. Eleven parameters were measured on each image 5 separate times by 2 independent observers using a computerized imaging program (NIH Image). Parameters evaluated were angle recess area, scleral thickness, peripheral corneal thickness, peripheral iris thickness, termination of Descemet's-iris distance, termination of Descemet's-ciliary process distance, angle opening distance, iris-lens contact, and the maximum length, width and area of the ciliary cleft. The coefficient of variation (CV) of each observer was used to assess intra-observer reproducibility. CV of less than 10% was considered indicative of good reproducibility. Inter-observer reproducibility was assessed by two way ANOVA. Image reproducibility: 5 images were sequentially obtained in the same location of each of 8 eyes of 4 dogs. The same parameters were measured once on each image, and image reproducibility was assessed by the coefficient of variation for each parameter. Results:The coefficient of variation was less than 10% for observer A for all parameters measured except ciliary cleft area (CV=11.63%). CV was greater than 10% for observer B for the angle recess area (18.51%) and ciliary cleft width (17.44%) and area (16.01%). Inter-observer reproducibility was lower, with significant differences (p<0.05) for 5 of the 11 parameters measured. Image reproducibility was generally lower, with highest CV for angle recess area (CV=33.01) when measurements from different images were combined, and lowest for maximum length of the ciliary cleft (CV=11.32). Conclusions: High resolution ultrasound measurements can be reproducibly obtained, especially if one experienced observer is used to measure alterable parameters. Consideration should be given, however, to differences attributable to variations in the captured images or between observers so as to avoid erroneous conclusions regarding significant differences among measured parameters.
This PDF is available to Subscribers Only