Purchase this article with an account.
A. Fink, C. Kassner, K. Saleh, W. Vilser, T. Riemer, E. Nagel, G. Henning; Technical Problems Which Should Be Considered During The Manual Measurement Of Retinal Vessel Diameters In Fundus Images . Invest. Ophthalmol. Vis. Sci. 2004;45(13):2801.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: There are many indications that the investigation of the retinal microcirculation via the vessel diameters (VD) not only gives new insights about the causes of retinal diseases. It also points out coherences between measurable parameters in the eye and internal medicine (e.g. ARIC study). But there are several technical problems which should be considered before retinal VD are measured in order to minimize their influence on the measured values and to improve the comparability of studies performed by different centres. Methods: The investigations were done with the manual measure mode of the software package VesselMap2 (Imedos; this software can also perform automatic diameter measurements). Diameter measurements were performed with different settings of graphic adapter and monitor by 4 different investigators. Image material were real retina images, computer generated vessel images, compressed (JPEG, different compression levels) retina images and compressed computer generated images. Results: We could determine different technical parameters influencing the outcome of the VD measurement. These are the settings of graphic adapter and monitor and image disturbances after image compression. The latter is important in terms of a new trend in retina diagnostics; images are taken in small departments and transmitted via network to an assessment centre whereby transmission of compressed images is common practice. Conclusion: The consideration of the technical problems during the manual measurement of retinal VD is essential to improve correctness and comparability of such studies. The minimum requirement is keeping the conditions stable over the whole study and their exact description. This is the only way to avoid fluctuations caused by unstable technical conditions. It is recommended to use one measuring system for the whole study or to use clearly defined and calibrated graphic adapter – monitor combinations. Automatic systems for measuring retinal VD are not disturbed by the described technical problems (except for image compression) because they work on the original image data.
This PDF is available to Subscribers Only