Detection of Retinal Pigment Epithelial Changes by Nonmydriatic Fundus Cameras

I read with great interest the recent article by Klein et al¹ regarding digital imaging of age-related macular degeneration (ARMD). This was a validation study of a 6.3-megapixel digital fundus camera in dark-adapted pupils and in pharmacologically dilated pupils vs the gold-standard 35-mm fundus camera, and the study looked at macular changes of ARMD. Klein and colleagues concluded with the detection of the different grades of ARMD being adequate with the digital camera, especially when the pupils were pharmacologically dilated. I congratulate the authors on an elegant and ingenious study. By recognizing the importance of the current high-resolution cameras and the advantage of digital photography, this study could represent a turning point in ARMD imaging. It is now recognized that a 6.3-megapixel digital camera captures details of the retina that are still detectable with great clarity even after significant magnification, a quality lost with the traditional 35-mm slide format and digital cameras of lower resolution.

Klein and colleagues confirm what other previous studies had suggested before, that is, the adequacy of high-resolution digital images in replacing classical 35-mm photography in detecting ARMD changes.² However, 1 limitation in the nonmydriatic arm of this series was the lack of stereopsis.¹ According to Klein and colleagues, this hindered the recognition of certain retinal pigment epithelial changes that were distinguished in the rest of the study.¹ This flaw was also described in other studies that used digital imaging of retinal diseases.³ While some reading groups have used “surrogate” markers, such as relying on hard exudate rings and visual acuity in the case of diabetic macular edema,³ to overcome such a limitation, most digital imaging systems have acquired techniques to obtain stereoscopic views of the retina. Even nonmydriatic imaging systems have developed 3-dimensional viewing, which has led to better diagnosis and better assessment of “elevations” in retina or choroid pathological abnormalities at the macula. Thus, this advantage was used by Bursell et al⁴ in evaluating diabetic macular edema, which led to the achievement of better specificity and sensitivity in relation to diagnosing clinically significant diabetic macular edema.⁴ Klein and colleagues were using a nonmydriatic digital camera of higher resolution than the cameras that Bursell and colleagues used. Probably adding to the high-resolution capacity of the camera, a 3-dimensional viewing of the maculae following the model used by Bursell and colleagues would have substantially improved detection of ARMD changes in the dark-adapted pupils. I thank the authors, and I would recommend an extension of the validation study using stereoscopic viewing with nonmydriatic high-resolution cameras.

Correspondence: Dr Salti, Department of Ophthalmology, Retina Service, American University of Beirut Medical Center, PO Box 110236, Riad El Solh 1107 2020, Beirut 110236, Lebanon (hs06@aub.edu.lb).