Magnification characteristics of fundus imaging systems

To compare the magnification properties of 11 different fundus cameras (including 1 stereo fundus camera), a Rodenstock (infrared) scanning laser ophthalmoscope (SLO), the Heidelberg Laser Tomographic Scanner (LTS), and the Heidelberg Retina Tomograph (HRT).

A cross-sectional study of the relationship between the true size of a fundus feature and its photographic-computer image in 14 different fundus imaging devices. This relationship was evaluated for each instrument using a model eye adjusted for axial ametropia between +11 diopter (D) and -14 D. To simulate refractive ametropia, the "crystalline lens" was removed to render the model eye aphakic, and the axial length was adjusted to give aphakic ametropia from emmetropia to +20 D.

A correction factor (p) was calculated for each instrument, which can be used in calculations for determining true retinal size.

The following were found to be of telecentric construction, Zeiss Oberkochen (WS240 Heidelberg), Zeiss Oberkochen (UK), Zeiss Oberkochen (Cologne), Nikon NF505, Kowa RCXV, SLO prototype (UK), LTS, and the HRT, and each exhibited a constant relationship between p and degree of ametropia of the model eye. The Canon CF6OU, Canon CF6OS, Canon CR4-45NM, Nidek 3-DX, Olympus GRCW, and Carl Zeiss Jena Retinophot were found not to be telecentric and exhibited a linear relationship between p and degree of ametropia of the model eye. For all instruments, p remained unchanged for axial and refractive ametropias of the same degree.

The study has shown that not all fundus imaging systems are telecentric, so the use of a single magnification correction value may not be appropriate. These findings have important implications for the way in which true retinal size calculations are performed. Examples are given to show how the tabulated values of correction factors can be used for both telecentric and nontelecentric cameras in image size calculations.