Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 May 1.
Published in final edited form as: Ophthalmol Retina. 2020 Aug 29;5(5):450–457. doi: 10.1016/j.oret.2020.08.017

Quantitative Comparison of Fundus Images by Two Ultra-Wide Field Fundus Cameras

Andrew Chen 1, Suveera Dang 2, Mina M Chung 1,3, Rajeev S Ramchandran 1, Angela P Bessette 1, David A DiLoreto 1,3, David M Kleinman 1, Jayanth Sridhar 4, Charles C Wykoff 5, Ajay E Kuriyan 1,3,6
PMCID: PMC8501635  NIHMSID: NIHMS1690920  PMID: 32866664

Abstract

Purpose:

To compare the relative retinal pixels and retinal area imaged using the Optos P200DTx and Zeiss Clarus 500 ultra-wide field (UWF) fundus cameras.

Design:

Single-center retrospective cross-sectional analysis

Subjects:

78 eyes of 46 patients

Methods:

Eyes were imaged with both the Optos P200DTx and Zeiss Clarus 500 UWF fundus cameras. The Optos P200DTx photos were single capture images without montage. Clarus images were two single capture images that were montaged, when possible. The relative pixels encompassing all visible retina in four quadrants centered on the fovea were measured. Measurements were normalized to the optic disc area for each image to account for pixel density and image size differences. Additionally, retinal area was measured in all four quadrants using Optos proprietary software Area-Measurement Tool and ImageJ software, using Optos P200DTx images and Zeiss Clarus 500 images that were registered to the Optos P200DTx images. Patients and technicians were asked for their preference between the machines. Imaging session times were recorded.

Main Outcome Measure:

Relative retinal pixels and retina area captured by each fundus camera.

Results:

When comparing all 78 images, the Optos P200DTx consistently captured more relative pixels in all four quadrants compared to the Zeiss Clarus 500: 93.8 vs. 75.0 superiorly (p < 0.001), 102.6 vs. 61.6 inferiorly (p < 0.001), 146.9 vs. 91.2 temporally (p < 0.001), and 167.1 vs. 127.8 nasally (p < 0.001). For the area calculation, 70 (89.7%) of the 78 images had successful registration of Zeiss Clarus 500 images to Optos P200DTx images. The Optos captured a larger retinal area in all four quadrants: 169.3 mm2 vs 131.6 mm2 superiorly (p < 0.001), 169.3 mm2 vs. 112.5 mm2 inferiorly (p < 0.001), 213.3 mm2 vs 143.6 mm2 temporally (p < 0.001), and 220.7 mm2 vs 178.8 mm2 nasally (p < 0.001). Eighteen (21.8%) of the 78 Zeiss Clarus 500 images were single capture images due to the patient being unable to take multiple images and were unable to be montaged. When comparing only the 52 (74.3%) of 70 Zeiss Clarus 500 images that were able to be montaged with the Optos P200DTx images, there was similarly significantly more relative pixel count and retina area imaged by the Optos P200DTx images. Among the 66 images with peripheral pathology, the Optos P200DTx captured findings not captured by the Zeiss Clarus 500 in 28 (42.4%) images, and the Zeiss Clarus 500 captured findings not captured by the Optos P200DTx in 1 (1.5%) image (P < 0.001). Among the 48 imaging sessions in which technicians responded with a preference, the Optos P200DTx was preferred for 28 imaging sessions (58%) compared to the Zeiss Clarus 500 for 20 imaging sessions (42%, P = 0.15). Among 44 patients who responded with a preference, 24 preferred the Optos P200DTx, 20 preferred the Zeiss Clarus 500 (P = 0.52). The average imaging session time was 4.6 minutes (SD: 3.0) for the Optos P200DTx and 5.3 minutes (SD: 3.1) for the Zeiss Clarus 500 (p=0.17).

Conclusion:

In the current study, the Optos P200DTx captured statistically significantly more retinal area in all four quadrants compared to the Zeiss Clarus 500. There was no statistically significant difference in patient or technician preference or image acquisition time between the two devices.

Keywords: fundus photography, ultra-wide field, Optos, Zeiss, Clarus

Introduction

Clinically significant pathology of the peripheral retina is often present in sight threatening disease. Diabetic retinopathy and other proliferative retinopathies, retinal vascular occlusions, vasculitis, retinal detachments, choroidal masses, and retinopathy of prematurity are prime examples. Retinal fundus photography is an important diagnostic modality in ophthalmology, and developments in this technology over the last century have greatly enhanced ophthalmologist’s ability to appreciate, document, and evaluate peripheral retinal pathology.1,2 In 1926, Nordensen and the Carl Zeiss Company introduced a fundus camera capable of a 20-degree field of view. Modern fundus cameras based on this original technology are capable of a 30–60-degree view. Documentation of more peripheral retina pathology with the fundus camera images traditionally required montaging images or a special contact lens. These steps demand time, patient cooperation, good pupillary dilation and a skilled technician, all of which limit the feasibility of obtaining quality photographs of the peripheral retina.

More recently, devices capable of imaging greater than 100-degrees, termed ultra-widefield (UWF), have become commercially available. One of the more common systems is the Optos P200DTx noncontact camera, which includes pseudo-color fundus photography, autofluoresence, fluorescein angiography, and indocyanine green angiography. The Optos P200DTx combines confocal scanning laser ophthalmoscopy (cSLO), which helps eliminate artifact, and an ellipsoid mirror to obtain up to 200-degrees of view without pupillary mydriasis. Limitations of this system include more constricted superior to inferior view, distorted periphery, artificial coloration of the retina, and lack of imaging anterior to the equator.1,3 Software has been developed to minimize the peripheral distortion.4 The Zeiss Clarus 500 system was commercially released in the U.S. in 2017 and has color fundus and autofluorescence capability. In 2019, the Zeiss Clarus 700, which includes fluorescein angiography (FA) capabilities with the same peripheral imaging technology was released. Both Clarus 500 and 700 utilize an imaging technique called Broad Line Fundus Imaging that is a hybrid of cSLO and traditional fundus photography, which can potentially provide higher resolution images with more accurate coloration of the fundus. A single image capture with this system obtains 133-degrees of view. A 200-degree of view is achieved by montaging images of temporal and nasal retina.

Efforts to expand care opportunities to patients with diabetic retinopathy (DR) through teleophthalmology have demonstrated the advantages of UWF fundus photography over conventional nonmydriatic fundus photography (NMFP) and shown substantial agreement to the gold standard-stereoscopic 7-field Early Treatment Diabetic Retinopathy photography.5,6 UWF fundus photography demonstrated increased identification of DR, reduced ungradable rates, decreased acquisition time, and increased retinal area measured compared to NMFP.5 Furthermore, more peripheral lesions are associated with more severe disease and highly predictive of DR progression.7 UWF may be more consistent across varying populations in the US than traditional nonmydriatic fundus photography.8

Comparisons between the amount of retina visualized by the Optos P200DTx and the Heidelberg Spectralis fluorescein angiography images have demonstrated that the Optos was able to image a larger amount of retina.9 One study compared differences in visible retina in non-montaged fundus photos using the Zeiss Clarus 500 and the Optos P200DTx cameras in 27 non-dilated diabetic patients and found a larger amount of retina imaged by the Optos device.10 However, the major limitations of this comparison were that the lack of mydriasis for both devices, which affects the quality of the image for UWF cameras, and that none of the Zeiss images were montaged, which precluded the ability of the device to capture beyond 133° of the retina.11 To date, there has been no direct comparison in the amount of retina visualized by of the fundus images produced by the Zeiss Clarus 500 and the Optos P200DTx in eyes with mydriasis and using montaged Zeiss Clarus 500 images. Comparison of the amount of retina imaged between the two cameras can provide useful information about the extent of peripheral imaging possible. In this study, we compared the amount of retina imaged by the Optos P200DTx and the Zeiss Clarus 500 fundus cameras using two methods: relative pixels and retinal area.

Methods

A single-center, retrospective, cross-sectional analysis was performed on fundus photographs from eyes undergoing imaging with both the Optos P200DTx (Optos PLC, Dunfermline, United Kingdom) and Zeiss Clarus 500 (Carl Zeiss Meditec AG, Jena, Germany) cameras during a six day product comparison and purchasing evaluation (January 25, 2018 – January 30, 2018) of both instruments at an academic retina clinic. The order of image capture was varied depending on camera availability. The Optos P200DTx photos were single capture images. The Zeiss Clarus 500 images were automatically-montaged images of two separate captures, whenever possible. In some instances, the technician was unable to take multiple images for reasons such as patient difficulties with fixation target or discomfort. All eyes underwent mydriasis with tropicamide 1% (Akorn Inc., Lake Forest, IL, USA) and phenylephrine 2.5% (Paragon Biotek Inc., Portland, OR, USA). Institutional review board approval was obtained, and the study adhered to the Declaration of Helsinki. Imaging session times were recorded excluding patients who underwent FA with Optos P200DTx from the time analysis. Patients and technicians were surveyed at the time of the evaluation for ease of use and comfort for each modality.

Adobe Photoshop CC (Adobe Inc. Photoshop CC version 20.0.1. San Jose, CA, USA) was used to calculate the pixel area for all four quadrants centered on the fovea as previously described (Figure 1).12 Retinal area obscured by peripheral artifacts were excluded from the analysis. Optic disc pixel area was obtained by identifying corresponding locations on the disc margin and demarcating the area with the rectangular selection tool. Relative pixel area was calculated by dividing the raw pixel count by the optic disc pixel count for each image to account for pixel density and image size differences between each machine.

Figure 1.

Figure 1.

Open in a new tab

Fovea-centered quadrants of ultra-widefield retinal images captured by the Zeiss Clarus 500 (Carl Zeiss Meditec AG, Jena, Germany) (A) and Optos P200DTx (Optos PLC, Dunfermline, United Kingdom) (B). The dotted lines demarcate the pixels used in the calculation of relative pixel areas.

In order to compare retina area captured between the two devices, the Zeiss Clarus 500 images were registered to the Optos P200DTx images. The image registration was completed by Optos engineers. This was done by using an affine registration based on intensity to get a broad alignment of the Zeiss Clarus 500 image to the corresponding Optos P200DTx images. Subsequently, rotating Gabor filters were used to find the vessels in both images. A non-affine registration based on mutual information, which used the input from the vessel maps to improve the registration and create fine-grained registration of Zeiss Clarus 500 image to corresponding Optos P200DTx image. The Zeiss Clarus 500 image was then placed in a 4000×4000 pixel plane that corresponded to the Digital Imaging and Communications in Medicine (DICOM) Supplement 173 standard for measuring widefield retinal image areas using Optos proprietary software Area-Measurement Tool and FIJI/ImageJ (ImageJ version 1.52i; US National Institutes of Health, Bethesda, MD, USA) software assuming an axial length of 24 mm.13 Calculation of retinal area imaged was performed by creating a binary mask in FIJI to delineate retinal quadrants centered on the fovea of the Optos P200DTx images and Zeiss Clarus 500 registered images. Then the masks were used to calculate pixel counts and pixel area for each quadrant with Optos proprietary software Area-Measurement Tool (Figure 2).

Figure 2.

Figure 2.

Open in a new tab

Zeiss Clarus 500 (Carl Zeiss Meditec AG, Jena, Germany) (A, C) and Optos P200DTx (Optos PLC, Dunfermline, United Kingdom) (B, D) retinal images after image registration to correct for differences in image resolution and peripheral distortion with fovea-centered quadrant masks to indicate regions of interest in white for area calculation.

Retinal pathology was evaluated for all sets of images with successful registration to determine the extent of lesions captured by both imaging systems. Each set was compared in a side-by-side fashion by two graders (AC and AEK). Peripheral artifacts were assessed and tallied. A subjective image quality score was assigned to each image based on the ability to assess retinal features from 0 to 4, with 0: having no retinal features identifiable, 1: having retinal features fully identifiable in <25% of imaged peripheral retina, 2: having retinal features fully identifiable in <50% of imaged peripheral retina, 3: having retinal features fully identifiable in <75% of imaged peripheral retina, and 4: having fully identifiable retinal features (no obscuration by artifacts).

Statistical analysis was performed using SPSS (IBM Corp. Released 2018. IBM SPSS Statistics for Windows, Version 24. Armonk, NY, USA). The Wilcoxon signed rank test was used to compare pixel count, retinal area, and imaging time. The Fisher’s exact test was used to compare differences in ability to detect retinal pathology, patient preferences, and technician preferences between the two devices.

Results

78 eyes of 46 patients were imaged with both the Optos P200DTx and Zeiss Clarus 500 and included in the study. Eighteen (25.7%) of the 78 Zeiss Clarus 500 images were single capture images due to the technicians’ inability to take multiple images for reasons such as patient difficulties with fixation target or discomfort (Figure 3). Eight (10.4%) of the 78 Zeiss Clarus 500 images were unable to be registered to their corresponding Optos P200DTx images, which was required to calculate retinal area with the Optos proprietary software Area-Measurement Tool. A total of 70 images were registered and used for retinal area calculation and comparison. The mean (SD) age for patients was 58.4 years (18.4), and the mean average pupil size was 7.0 mm (1.4).

Figure 3.

Figure 3.

Open in a new tab

Single-capture Zeiss Clarus 500 (Carl Zeiss Meditec AG, Jena, Germany) (A) versus single-capture Optos P200Dtx (Optos PLC, Dunfermline, United Kingdom) (B) images.

Relative Pixel Count Analysis

The Optos P200DTx captured a larger total relative pixel count than the Zeiss Clarus 500, 510.4 vs 355.6 (P <0.001). The Optos P200DTx also consistently captured a larger relative pixel count in all four quadrants compared to the Zeiss Clarus 500 (Table 1): 93.8 vs. 75.0 superiorly (P < 0.001), 102.6 vs. 61.6 inferiorly (P < 0.001), 146.9 vs. 91.2 temporally (P < 0.001), and 167.1 vs. 127.8 nasally (P < 0.001). Even when comparing only the subset of Zeiss Clarus 500 images that were able to be montaged (n = 61) with the corresponding Optos P200DTx images, the Optos P200DTx consistently captured a larger relative pixel count in all four quadrants compared to the Zeiss Clarus 500 (Table 1): 89.0 vs. 75.5 superiorly (P = 0.001), 96.3 vs. 62.1 inferiorly (P < 0.01), 138.4 vs. 99.0 temporally (P < 0.001), and 160.1 vs. 138.0 nasally (P = 0.02).

Table 1.

Comparison of relative pixel count between the Optos P200DTx and the Zeiss Clarus 500. Absolute pixel counts for each quadrant were normalized by optic nerve pixel count to account for pixel density and size differences. Ratio represents the relative pixel count of Optos P200DTx divided by Zeiss Clarus 500.

All Images (n=78) Montaged Images (n=61)
Quadrant Optos Zeiss Ratio P Value Optos Zeiss Ratio P Value
Superior 92.4 74.9 1.2 <0.001 89.0 75.5 1.2 0.001
Inferior 100.7 61.3 1.6 <0.001 96.3 62.1 1.6 <0.001
Temporal 145.0 91.2 1.6 <0.001 138.4 99.0 1.4 <0.001
Nasal 165.2 128.3 1.3 <0.001 160.1 138.0 1.2 0.02
Total 510.4 355.6 1.4 <0.001 483.8 374.7 1.3 0.01
Open in a new tab

Retina Area Analysis

Seventy (89.7%) of 78 Zeiss Clarus 500 fundus images were able to be registered, of which 18 of 70 (25.7%) were non-montaged images. Optos P200DTx captured a larger total retinal area than the Zeiss Clarus 500, 765.6 mm2 vs 566.5 mm2 (P < 0.001). Retinal area captured of Optos P200DTx versus registered Zeiss Clarus 500 images are listed in Table 2. The Optos P200DTx captured a larger retinal area in all four quadrants: 169.3 mm2 vs 131.6 mm2 superiorly (P < 0.001), 162.4 mm2 vs. 112.5 mm2 inferiorly (P < 0.001), 213.3 mm2 vs 143.6 mm2 temporally (P < 0.001), and 220.7 mm2 vs 178.8 mm2 nasally (P < 0.001). Total retinal area captured of Optos P200DTx versus registered Zeiss Clarus 500 image were 765.6 mm2 versus 566.5 mm2 respectively (P < 0.001). Even when comparing only the subset of Zeiss Clarus 500 images that were able to be montaged (n = 52) with corresponding Optos P200DTx images, the Optos P200Tx consistently captured a larger retinal area in all four quadrants compared to the Zeiss Clarus 500 (Table 2): 168.8 mm2 vs. 136.1 mm2 superiorly (P < 0.001), 161.6 mm2 vs. 117.8 mm2 inferiorly (P < 0.001), 213.0 mm2 vs. 158.0 mm2 temporally (P < 0.001), and 219.8 mm2 vs. 196.1 mm2 nasally (P = 0.001). Total retinal area for registered images from the Optos P200DTx and the Zeiss Clarus 500, were 763.2 mm2 and 608.1 mm2, respectively (P < 0.001).

Table 2.

Comparison of Registered Retinal Area between the Optos P200DTx and Zeiss Clarus 500

All Registered Images (n = 70)
 All Registered Montaged Images (n = 52)
Quadrant Optos (mm2) Zeiss (mm~) Ratio P Value Optos (mm~) Zeiss (mm2) Ratio P Value
Superior 169.3 131.6 1.3 <0.001 168.8 136.1 1.2 <0.001
Inferior 162.4 112.5 1.4 <0.001 161.6 117.8 1.4 <0.001
Temporal 213.3 143.6 1.5 <0.001 213.0 158.0 1.3 <0.001
Nasal 220.7 178.8 1.2 <0.001 219.8 196.1 1.1 0.001
Total 765.6 566.5 1.4 <0.001 763.2 608.1 1.3 <0.001
Open in a new tab

Zeiss Clarus 500 images were registered to Optos P200DTx images to standardize resolution, centration, and peripheral distortion. Ratio represents the retina area imaged by Optos P200DTx divided by that of the Zeiss Clarus 500.

Retinal Pathology

The ability to detect retinal pathology was compared between the 70 sets of registered images. Differences in the area or extent of the pathology was not included in this analysis, only that ability to visualize if the pathology was present in the images. Of the 70 images, 26 eyes were normal and were not included in the analysis. 6 of 44 (13.6%) eyes had multiple findings in the macula, and 10 of 56 (17.9%) eyes had multiple findings inthe periphery. There were 44 eyes with 50 counts of pathology within the posterior pole. Pathology included DR 15 (30.0%), age-related macular degeneration (AMD) 11 (22.0%), epiretinal membrane 8 (16.0%), chorioretinal scar 4 (8.0%), vascular occlusion 3 (6.0%), posterior uveitis 3 (6.0%), history of retinal detachment 2 (4.0%), peripapillary atrophy 2 (4.0%), choroidal neovascular membrane 1 (2.0%), retinoschisis 1 (2.0%). All 44 of 44 eyes with posterior pole pathology were detected by both the Optos P200DTx and Zeiss Clarus 500. There were 56 eyes with 66 counts of pathology peripheral to the posterior pole, of which pathology included: peripheral degeneration 28 (42.4%), DR 13 (19.7%), peripheral drusen 10 (15.2%), treated retinal detachment or tear 5 (7.6%), posterior uveitis lesions 4 (6.1%), nevus 3 (4.5%), and retinoschisis 1 (1.5%). The Optos P200DTx captured pathology not captured by the Zeiss Clarus 500 in 28 (42.4%) cases, and the Zeiss Clarus 500 captured pathology not captured by Optos P200DTx in 1 case (1.5%, P < 0.001). The additional pathology that was detected by the Optos P200DTx was due to increased area of retina imaged. The additional pathology that was detected by the Zeiss Clarus 500 was a temporal chorioretinal scar captured only by the Zeiss Clarus 500, due to due to the peripheral location of the lesion. The additional peripheral findings were only deemed clinically significant in 8 (14.3%) eyes (diagnoses: history of detachments or tears 3 (4.5%), DR 2 (3.0%), peripheral lattice degeneration 1 (1.5%), vascular occlusion 1 (1.5%), and posterior uveitis 1 (1.5%)), all of which were captured by Optos P200DTx. Of note, 16 of the 70 (22.9%) Zeiss Clarus 500 images were unable to be montaged. When only comparing the montaged Zeiss Clarus 500 images, the Optos P200DTx captured pathology not captured by the Zeiss Clarus 500 in 24 of 48 (50.0%) cases of peripheral findings, and the Zeiss Clarus 500 captured pathology not captured by Optos P200DTx in 1 case (2.1%, P < 0.001).

Image Artifacts and Image Quality

Lids or lashes were noted in 69 of 70 (98.6%) Optos P200DTx images, most commonly in the inferior quadrant (67, 95.7%) versus superior quadrant (57, 81.4%). Lids or lashes were noted in 27 of 70 (38.6%) Zeiss Clarus 500 images, most commonly in the inferior quadrant (21, 30.0%) versus in the superior quadrant (9, 12.9%). Peripheral artifacts from the Clarus montaging algorithm was observed in 10 (14.3.9%) cases. Image quality for the Optos P200DTx and Zeiss Clarus 500 were graded 2.7 ± 0.5 and 2.9 ± 0.9 respectively (P = 0.26).

Imaging Time, Preference, and Image Quality Comparison

Average time for Optos P200DTx image acquisition was 4.6 minutes (3.0) for 39 patients. Average time for Zeiss Clarus 500 image acquisition was 5.2 minutes (3.0) for 44 patients (P = 0.4). Among the 48 imaging sessions in which technicians responded with a preference, the Optos P200DTx was preferred for 28 imaging sessions (58%) compared to the Zeiss Clarus 500 for 20 imaging sessions (42%, P = 0.15). Among 44 patients who responded with a preference, 24 preferred the Optos P200DTx, 20 preferred the Zeiss Clarus 500 (P = 0.52).

Discussion

Evaluation of the peripheral retina is critical for optimal patient management in ophthalmology. Retinal diseases such as diabetic retinopathy, retinal detachments, retinal vein occlusions, and uveitis are major causes of vision loss, and good outcomes can depend on detailed and accurate physician impressions of the retinal periphery. UWF imaging modalities have been shown to reveal significantly more pathology compared to conventional fundus photography.14,15 UWF FA reveals more pathology when compared to seven standard field imaging for diabetic retinopathy.9,16 For uveitis, UWF FA can detect increased pathology that would otherwise be missed on conventional angiography.17–19 Currently commercially available UWF fundus cameras include the Optos P200DTx, Optos P200TE, Optos P200T, and Zeiss Clarus 500 and 700.

The Optos cameras combine cSLO and an ellipsoid mirror to block scattered light from media opacities and to image up to a reported 200 internal degrees of retina in the horizontal axis.20 Superimposing images obtained from two different colored lasers produces a semi-realistic color image. The system can obtain UWF images of the retina without mydriasis, but mydriasis can improve image quality.11

In comparison, the Zeiss Clarus 500 and 700 systems utilize broad line fundus imaging, which is a technique patented by Zeiss that is a hybrid of both cSLO and traditional fundus photography and utilizes line scanning illumination with light-emitting diodes and an aperture confocal to the illumination.21 A single image captures 133 internal degrees. With a montage of two images, the Zeiss Clarus can reportedly capture 200 internal degrees of retina.22 One reported advantage of the Clarus system is the color accuracy of the color fundus photos acquired22, although no studies have been conducted to determine if the difference in coloration results in clinically meaningful differences. At the time of the image acquisition, the Clarus system was only able to acquire color fundus, fundus autofluorescence, and fluorescein angiography photos. The ability to perform indocyanine green angiography is anticipated to be commercially available at a future date.

There have been few direct comparisons between the UWF imaging devices. In a direct comparison of UWF FA imaging of Optos P200Tx and Heidelberg Spectralis HRA + OCT, Witmer et al. found that Optos captured more total, temporal and nasal retinal surface pixel area. Overall, Spectralis captured more superior and inferior retinal vasculature, though the number of pixels in these quadrants were not statistically different.12 However, the Spectralis device cannot obtain color fundus images. Hirano et al. compared peripheral retina imaged in a single capture of the Optos P200DTx and the Zeiss Clarus 500 color fundus image for assessment of diabetic retinopathy in 46 eyes of 25 patients. They report an average of 465 disc areas captured by the Optos P200DTx versus 243 disc areas by the Zeiss Clarus 500.10 Because a single capture of the Zeiss Clarus 500 was never intended to image the same extent of peripheral retina as the Optos P200DTx, it is difficult to use these results to evaluate the ability of each device to image the peripheral retina. Furthermore, the images were taken of undilated eyes, which can affect image quality.10

Therefore, we compared the extent of retina imaged by leading UWF systems in dilated eyes, using single shot images captured on the Optos P200DTx and montaged images taken on the Zeiss Clarus 500. On average, the Optos P200DTx was able to capture significantly more relative retina pixels and retinal area in each individual quadrant, as well as all four quadrants combined, compared to the Zeiss Clarus 500. Acquisition time, patient preference, and technician preference did not differ between the two imaging devices.

All UWF camera images have a degree of artifact as a result of projection from the 3D spherical retinal plane onto a 2D image. This results in distortion such that a single peripheral pixel represents a smaller surface area than a single posterior pixel, causing an over-representation of the retinal periphery, compared to the posterior pole, in the 2D image.13 There are different methodologies used to create the 2D image; some may attempt to preserve relative anatomical structure, while others may attempt to preserve relative size. The DICOM Committee attempts to address this issue and created a standard for projection of non-flat images onto a 2D space. However, different companies may opt for different methodologies for creation of a 2D image. Consequently, direct area quantification of retinal pixel area may not always be comparable between different parts of the retina or between different machines.4,13 Therefore, we felt that performing only a relative pixel analysis when comparing images obtained from the two devices may not provide the ideal comparison. To address this issue, we registered the fundus photos of the same eye taken on the Zeiss Clarus 500 to the Optos P200DTx, based on vasculature landmarks. After registering the images, we were able to use Optos proprietary software Area-Measurement Tool and FIJI/ImageJ software to measure the retinal area captured by the two devices.

Though evaluation of the ability to detect retinal pathology was performed, and the Optos P200DTx captured a statistically significantly larger proportion of peripheral retinal pathology not seen with the Zeiss Clarus 500, the majority of these findings were not clinically significant. The order in which the images were taken was varied based on camera availability, and the lack of true randomization could have introduced unintentional bias. Whether or not fatigue among patients who underwent Clarus imaging second affected capture quality was not able to be analyzed. Evaluation of retinal pathology was performed on a side-by-side basis, which could have introduced bias. Additionally, differences in extent of lesions were observed, but not included in this analysis. Further work will need to be performed to determine if the differences in peripheral retinal captured would affect clinical management.

One limitation of this study was the inability to montage 25.7% of Zeiss Clarus images, primarily due to patient discomfort from the flash. Therefore, we performed the comparison analysis for all images and only montaged images and found similar outcomes. We chose to include the comparison of the non-montaged Zeiss Clarus images in our analysis to provide a “real world” assessment. 10.4% sets of images could not be registered due limited extent of anatomical landmarks for the registration software to use that resulted in poor quality registration of images (e.g. media opacities, poorly dilating pupils). Another limitation is that we did not use montaged Optos P200DTx images, which would have potentially increased the extent of retina imaged by that device. Although this may result in underestimating the capability of the device, we sought to mirror the image acquisition technique that is most widely used, which is a single capture image. Despite this limitation, the OptosP200DTx images captured more retina overall than the Zeiss Clarus 500 in our study. While Zeiss Clarus 500 images were registered to Optos P200DTx images for area calculations using Optos software, we were unable to register the Optos P200DTx images to the Zeiss Clarus 500 images because the Zeiss Clarus 500 images were stored as JPG files. JPG files are not able to be used with the Clarus Review Software or Clarus instrument. Another limitation is that this study was not designed to assess the ability to detect differences in pathology, impact of peripheral distortion, resolution of images, or coloration of fundus photos. Although differences in these factors between the imaging modalities may be important to retina specialists, such analysis was outside the scope of the study which was intended to only compare the amount of retina visible between the two devices. Furthermore, this was small group of patients who were relatively young with good pupillary dilation. We utilized subjective grading scale to assess differences in image quality based on artifacts, but no significant difference was noted between the two cameras. Another limitation of the study is the inability to blind the grader/assesser of images to the camera used to take the image due to distinct differences in the appearance of the images from each camera. Lastly, no significant difference was observed for image acquisition time or patient comfort, but the result possibly reflects the small sample size and lack of power. Technicians did not have prior experience with either device. A strength of the study is that the clinic did not have either device prior to this study providing the opportunity to assess technician and patient preference without introducing any bias from familiarity with the machines.

In this study, the Optos P200DTx imaged a larger amount of retina (by relative pixels and area) than the Zeiss Clarus 500. Future studies are needed to assess if the differences in retina imaged are clinically significant in different disease states.

Supplementary Material

1

Precis.

We found the Optos P200DTx captured more peripheral retinal (pixels and area) in all four quadrants compared to the Zeiss Clarus 500. There was no difference in patient or technician preference or image acquisition time between the two devices.

Acknowledgments

Funding: Unrestricted grant from Research to Prevent Blindness (Flaum Eye Institute, University of Rochester Medical Center) and NIH P30 EY001319

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Witmer MT, Kiss S. Wide-field Imaging of the Retina. Surv Ophthalmol 2013;58:143–154. [DOI] [PubMed] [Google Scholar]
  • 2.Nagiel A, Lalane RA, Sadda SR, Schwartz SD. Ultra-Widefield Fundus Imaging: A Review of Clinical Applications and Future Trends. Retina Phila Pa 2016;36:660–678. [DOI] [PubMed] [Google Scholar]
  • 3.Silva PS, Cavallerano JD, Sun JK, et al. Nonmydriatic ultrawide field retinal imaging compared with dilated standard 7-field 35-mm photography and retinal specialist examination for evaluation of diabetic retinopathy. Am J Ophthalmol 2012;154:549–559.e2. [DOI] [PubMed] [Google Scholar]
  • 4.Singer M, Sagong M, van Hemert J, et al. Ultra-widefield Imaging of the Peripheral Retinal Vasculature in Normal Subjects. Ophthalmology 2016;123:1053–1059. [DOI] [PubMed] [Google Scholar]
  • 5.Silva PS, Horton MB, Clary D, et al. Identification of Diabetic Retinopathy and Ungradable Image Rate with Ultrawide Field Imaging in a National Teleophthalmology Program. Ophthalmology 2016;123:1360–1367. [DOI] [PubMed] [Google Scholar]
  • 6.Kernt M, Hadi I, Pinter F, et al. Assessment of Diabetic Retinopathy Using Nonmydriatic Ultra-Widefield Scanning Laser Ophthalmoscopy (Optomap) Compared With ETDRS 7-Field Stereo Photography. Diabetes Care 2012;35:2459–2463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Silva PS, Cavallerano JD, Haddad NMN, et al. Peripheral Lesions Identified on Ultrawide Field Imaging Predict Increased Risk of Diabetic Retinopathy Progression over 4 Years. Ophthalmology 2015;122:949–956. [DOI] [PubMed] [Google Scholar]
  • 8.Silva PS, Cavallerano JD, Tolls D, et al. Potential Efficiency Benefits of Nonmydriatic Ultrawide Field Retinal Imaging in an Ocular Telehealth Diabetic Retinopathy Program. Diabetes Care 2014;37:50–55. [DOI] [PubMed] [Google Scholar]
  • 9.Wessel MM, Aaker GD, Parlitsis G, et al. Ultra-wide-field angiography improves the detection and classification of diabetic retinopathy. Retina Phila Pa 2012;32:785–791. [DOI] [PubMed] [Google Scholar]
  • 10.Hirano T, Imai A, Kasamatsu H, et al. Assessment of diabetic retinopathy using two ultra-wide-field fundus imaging systems, the Clarus® and OptosTM systems. BMC Ophthalmol 2018;18:332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Rasmussen ML, Broe R, Frydkjaer-Olsen U, et al. Comparison between Early Treatment Diabetic Retinopathy Study 7-field retinal photos and non-mydriatic, mydriatic and mydriatic steered widefield scanning laser ophthalmoscopy for assessment of diabetic retinopathy. J Diabetes Complications 2015;29:99–104. [DOI] [PubMed] [Google Scholar]
  • 12.Witmer MT, Parlitsis G, Patel S, Kiss S. Comparison of ultra-widefield fluorescein angiography with the Heidelberg Spectralis(®) noncontact ultra-widefield module versus the Optos(®) Optomap(®). Clin Ophthalmol Auckl NZ 2013;7:389–394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Croft DE, van Hemert J, Wykoff CC, et al. Precise montaging and metric quantification of retinal surface area from ultra-widefield fundus photography and fluorescein angiography. Ophthalmic Surg Lasers Imaging Retina 2014;45:312–317. [DOI] [PubMed] [Google Scholar]
  • 14.Lengyel I, Csutak A, Florea D, et al. A Population-Based Ultra-Widefield Digital Image Grading Study for Age-Related Macular Degeneration-Like Lesions at the Peripheral Retina. Ophthalmology 2015;122:1340–1347. [DOI] [PubMed] [Google Scholar]
  • 15.Domalpally A, Clemons TE, Danis RP, et al. Peripheral Retinal Changes Associated with Age-Related Macular Degeneration in the Age-Related Eye Disease Study 2: Age-Related Eye Disease Study 2 Report Number 12 by the Age-Related Eye Disease Study 2 Optos PEripheral RetinA (OPERA) Study Research Group. Ophthalmology 2017;124:479–487. [DOI] [PubMed] [Google Scholar]
  • 16.Price LD, Au S, Chong NV. Optomap ultrawide field imaging identifies additional retinal abnormalities in patients with diabetic retinopathy. Clin Ophthalmol Auckl NZ 2015;9:527–531. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Karampelas M, Sim DA, Chu C, et al. Quantitative Analysis of Peripheral Vasculitis, Ischemia, and Vascular Leakage in Uveitis Using Ultra-Widefield Fluorescein Angiography. Am J Ophthalmol 2015;159:1161–1168.e1. [DOI] [PubMed] [Google Scholar]
  • 18.Pecen PE, Petro KF, Baynes K, et al. Peripheral Findings and Retinal Vascular Leakage on Ultra-Widefield Fluorescein Angiography in Patients with Uveitis. Ophthalmol Retina 2017;1:428–434. [DOI] [PubMed] [Google Scholar]
  • 19.Campbell JP, Leder HA, Sepah YJ, et al. Wide-field retinal imaging in the management of noninfectious posterior uveitis. Am J Ophthalmol 2012;154:908–911.e2. [DOI] [PubMed] [Google Scholar]
  • 20.Kernt M, Schaller UC, Stumpf C, et al. Choroidal pigmented lesions imaged by ultra-wide-field scanning laser ophthalmoscopy with two laser wavelengths (Optomap). Clin Ophthalmol Auckl NZ 2010;4:829–836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Bublitz D, Everett MJ, FARKAS C, et al. Systems and methods for broad line fundus imaging. 2014. Available at: https://patents.google.com/patent/WO2014140256A2/en [Accessed April 25, 2020].
  • 22.“CLARUS 500”. Zeiss. Available at: https://www.zeiss.com/meditec/int/product-portfolio/retinal-cameras/clarus-500.html [Accessed February 2, 2020].

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1690920-supplement-1.docx (16.4KB, docx)

RESOURCES