• S Kumar, K Yogesan, and I J Constable.
• Centre of Excellence in e-Medicine, University of Western Australia, Perth, Australia. sajeesh@cyllene.uwa.edu.au
• Eye (Lond). 2009 Mar 1; 23 (3): 652-60.

PurposeThis study compared digital images from a portable slit-lamp camera with 35 mm slit-lamp photographs and traditional ophthalmic assessments in anterior segment disorder's detection.MethodsA total of 196 patients (392 eyes) were recruited from an anterior segment ophthalmology clinic. Each patient underwent an examination by an anterior segment ophthalmologist. Two to three standardized views of 640 x 480 pixels digital images (portable digital slit-lamp camera) and 35 mm photographic slides (Zeiss slit-lamp camera) were taken after the examination. The same ophthalmologist reviewed these images in a masked fashion. Two other masked graders also assessed the digital images. The presence or absence of 33 specific findings was noted at each examination.ResultsDigital images showed moderate to excellent agreement to clinical findings (kappa 0.45-0.82) in areas other than lid pathologies. Lens findings from digital images had moderate to good agreement with the clinical gold standard (unweighted kappa 0.43-0.65, sensitivity 59-77%, specificity 86-94%). Gross cornea signs were well detected with digital images, (kappa 0.72-0.85, sensitivity 67-100%, specificity 98-99). More subtle corneal, conjunctival and lid abnormalities were not identified well. The statistical figures were very similar to the above-mentioned figures when the 35-mm film results were compared to clinical diagnoses. The two image formats showed better agreement when compared to each other than when either is compared with clinical findings.ConclusionDiagnoses using digital slit-lamp images were comparable to diagnosis using 35 mm photographic slides for some anterior segment abnormalities.

23 keywords...

hide…