Purpose:

To compare cataract images obtained using portable slit-lamp devices with those taken from the standard slit-lamp built-in camera.

Design:

Prospective, single-center, comparative digital imaging validation study.

Methods:

A total of 198 slit-lamp images were taken from 66 patients with cataracts using a smartphone slit-lamp prototype (“MIDAS”), a tablet with a portable slit-lamp (“I.C.P.”), and a slit-lamp with a custom-mounted digital camera (“ASC”). An ophthalmologist, an optometrist, and a family medicine resident independently graded the images according to Nuclear Opalescence (NO) and Nuclear Colour (NC) of the Lens Opacities Classification System III. The intraclass correlation (ICC) for the devices was calculated. Graders’ confidence and subjective assessment of image quality were assessed through a 5-point Likert scale questionnaire.

Results:

For NO, ICC was fair for MIDAS (0.619, 95% confidence interval [95% CI] = 0.426 to 0.754) and I.C.P. (0.585, 95% CI = 0.373 to 0.733) but poor for ASC (0.394, 95% CI = 0.088 to 0.609). For NC, ICC was excellent for all (MIDAS, 0.837, 95% CI = 0.755 to 0.895; I.C.P., 0.851, 95% CI = 0.775 to 0.904; ASC, 0.892, 95% CI = 0.837 to 0.930). Grader confidence was the highest for ASC images (85.5%), followed by I.C.P. (69.1%) and MIDAS (65.6%). Subjective assessment of image quality was the highest for ASC images (96.5%), followed by I.C.P. (77.3%) and MIDAS (74.2%).

Conclusions:

Portable slit-lamp devices had excellent ICC for assessing NC and fair ICC for assessing NO. Graders expressed greater confidence and higher perceived image quality for the standard slit-lamp built-in camera.