A Standardized Toolkit to Compare Next-Generation Fundus Cameras

Abstract

Purpose : Advances in technology have enabled the proliferation of portable fundus cameras that are now routinely used in clinical care. However, these vary dramatically in the quality of their output, and selecting the appropriate device for a specific clinical need remains a subjective endeavor. The purpose of our investigation was to develop a low-cost, standardized toolkit to objectively test the output characteristics of any given fundus camera.

Methods : For this evaluation of technology, a 3D printable model eye was constructed. The model was designed to allow for interchangeable test “retinae” intended to evaluate clinically relevant imaging characteristics: area imaged, color fidelity, resolution, distortion, and contrast. A smartphone with a 3D printed lens module and a commercially available hand-held fundus camera were used to image the test plates through the model eye. Resultant images were analyzed using National Institutes of Health ImageJ image analysis software. Sample analysis and comparison are included.

Results : The smartphone could visualize 130 mm² compared to 170 mm² with the commercial device, approximating a 60 degree fundus photo (115 mm²). Grid imaging readily identified pincushion (smartphone) vs. barrel (commercial) distortion. On RGB analysis, both showed poor color fidelity, with the smartphone tending to diffuse color (deltaE₇₆ = 24.1) and the commercial device shifting to extreme ends of the color spectrum (deltaE₇₆ = 59.9). Smartphone central resolution was 16 line pairs (lp)/mm vs. commercial 32 lp/mm (Figure 1); both failed to meet the requirement of 60 lp/mm required of clinical grade fundus cameras. Surface plot analysis demonstrated superior contrast sensitivity for the smartphone compared to the commercial device.

Conclusions : The approach described permits rapid and intuitive assessment of any fundus camera. This allows clinicians to more objectively understand the strengths and weaknesses of their tools and appropriately match devices with their desired clinical task. For example, a glaucoma screening program may opt to minimize distortion of cup-to-disk ratio whereas documentation of subretinal fluid may necessitate absolute color fidelity. More studies are needed to evaluate the reproducibility of test images and identify clinically meaningful differences.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

Figure 1. Standardized resolution plate imaged by smartphone (A) and commercial device (B)