The use of ultra-widefield (UWF) imaging has been shown to provide significant clinical value in the examination and diagnosis of retinal pathology due to improved ability to capture the peripheral retina. Historically, imaging of the peripheral retina has been limited with conventional machines. However, UWF imaging has made it possible to resolve more than 80% of the fundus with a single picture, allowing clinicians access to an improved view of peripheral clinical features that can aid in disease detection.1 Additionally, the ability to create a montage allows for almost the entire retinal surface to be analyzed.2
As fundus cameras with up to 50° field of view appeared, they became widely known as widefield imaging technology and were used in clinical practice for their ability to capture some aspects of a retina's periphery. More recently, technology has advanced with the conception of UWF fundus, which indicates a greater than 50° field of view through either a single image or automated imaging montage. The conception of UWF has led to numerous fundus imaging observations that demonstrate the presence or development of substantial pathology in the peripheral retina.3–8 Wessel et al. reported that UWF images showed 3.9-times more nonperfusion area (NPA) and 1.9-times more neovascularization than traditional Early Treatment Diabetic Retinopathy Study (ETDRS) seven-field images.3 Silva at al. reported that the retinal NPA at the midperiphery and far periphery detected on UWF images altogether accounted for approximately 86% of the total ischemic area.9
Several imaging platforms purport to provide a UWF view of the retina. Two examples include a widefield confocal scanning laser ophthalmoscope (WSLO) and a widefield broad line fundus (WBLF) camera. The first system utilizes a WSLO in combination with an ellipsoid mirror to produce images with approximately 200 internal degrees of view and create a two-color image of the retina using broad-spectrum red and green wavelengths.10 The 633-nm red and 532-nm green lasers penetrate the retina to different depths, including below the retinal pigment epithelium, and manipulation of these lasers by the ellipsoid mirrors allows maximizing the retinal area captured. This allows for simultaneous evaluation of the peripheral and central retina without significant eye-steering.
The WBLF system scans the retina with a broad rectangle of light detected by a monochromatic camera, and sequentially illuminates the retina with broad-spectrum red, green, and blue light-emitting diodes (LEDs) to produce true color images with 7 μm resolution. A single image can capture up to 133° of the retina, and multiple image captures (between two and six) can then be montaged to achieve 200° of view.11,12
Since the UWF technology is relatively new, a limitation presently is that widefield (WF) and UWF terminology have often been used interchangeably in the literature. Recently, the International Widefield Imaging Study Group recommended the classification and guidelines based on the nomenclature of imaging.13 The group suggested that instead of characterizing photos by the degrees, the terminology used to describe photos should incorporate anatomical landmarks within the retina, which is more clinically useful to measure and better reflects value for an ophthalmologist's clinical exam. Their definition of a WF image includes a single image centered on the fovea that captures the retina up to and including the posterior border of the vortex vein ampullae in all four quadrants. Comparatively, an UWF image was defined as capturing the retina in the far periphery in all four quadrants beyond the vortex vein ampulla in a single image capture of the retina. The purpose of this study was to evaluate two different technologies in terms of their WF capabilities in routine clinical practice.
Patients and Methods
After Cleveland Clinic Investigational Review Board approval was obtained, a retrospective chart review was performed in patients within routine clinical practice who were imaged at Cole Eye Institute, Cleveland, Ohio, with a WSLO (Optos P200DTx/California; Optos plc, Dunfermline, United Kingdom), and a WBLF imaging ophthalmoscope with both single and auto-montage images (Clarus 500; Carl Zeiss Meditec, Dublin, CA). Patients were imaged between June and July 2018 by multiple experienced ophthalmic technicians. All study-related procedures were performed in accordance with good clinical practice (International Conference on Harmonization of Technical Requirements of Pharmaceuticals for Human Use E6).
A total of 48 patients 18 years of age or older were identified. All participants had a diagnosis of a retinal disease in at least one eye and had the ability to adequately fixate to allow for high-quality images. Seven eyes with previous laser treatment or with poor photographs due to significant media opacities were excluded.
The primary outcome was the number of quadrants visible in the fundus image. “Visible” was defined as the quadrant having a defined vortex ampullae, thereby incorporating the International Wide-Field Imaging Study Group definition. To count vortex ampullae, images were viewed in green-free channel (red reflectance only) to accentuate the choroidal layer. Two trained masked reviewers (TC and NC) analyzed the images with a tertiary grader (RS) for adjudication. In some anatomical presentations, where multiple ampullae occurred in the same quadrant, they were recorded as a singular positive finding (Figure 1).
Example images with visible vortex ampullae circled. WBLF = widefield broad line fundus; WSLO = widefield confocal scanning laser ophthalmoscope
The secondary analysis included a quantitative assessment of quadrant size in millimeters squared (mm2). Images were transformed to stereographic projection images using proprietary prototype software available from the manufacturer. This projection technique was accomplished by ray tracing every pixel through a combined optical model of the Optos 200Tx and a Navarro UWF model eye with an axial length of 24 mm.14 This optical model represented the projection used by the Optos 200Tx scanning laser ophthalmoscopy platform to create the two-dimensional (2-D) Optomap (Optos, Dunfermline, United Kingdom). The area within each of four visualized quadrants was calculated using ImageJ (Version 1.52a; NIH, Bethesda, MD) and images were standardized to a 40 × 40 pixel size. Each image was split into four quadrants centered on the fovea, and the area of visible retina was outlined with a free-style line ImageJ tool demarcating only the visible areas of the retina (Figure 2). All artifacts or obscured areas of the retina were excluded. This area was then colored with the background removed and the area in millimeters squared (mm2) was measured using a customized measurement tool. The measurement tool provided output on the total area visualized (entire visible retina), as well as the area within each quadrant (superior, inferior, nasal, and temporal). The disc area was also marked and measured separately and used as a control to ensure images were standardized.
The outlined area of visible retina. (A) Widefield broad line fundus (WBLF) single image. (B) WBLF montage image. (C) Widefield confocal scanning laser ophthalmoscope image.
The ability to discern differences in clinical grading between the cameras was performed by having two masked readers (KT and CG) with expertise in retinal diseases read each image. A tertiary grader (RS) was incorporated for adjudication. All readers were masked to all clinical characteristics and other diagnostic information (eg, optical coherence tomography, fundus autofluorescence, etc.). Both technologies were evaluated in relation to the extension and severity of individual retinal disease.
Descriptive summaries were presented as means and standard deviations. A Kruskal-Wallis test was used to compare means of area that were obtained within each quadrant and total area between all technologies. An unpaired t-test was used to compare WSLO and WBLF images. A significance level of .05 was assumed for all tests. All statistical analysis was performed using Minitab software (Minitab, State College, PA).
A total of 48 patients were identified and 41 patients (65 eyes) were included in the study. Of the 41 patients, Table 1 demonstrates the clinical diagnoses of the patients included. The WSLO captured 116 of the possible 260 vortex ampullae (45%), whereas the WBLF single image captured eight of 260 vortex ampullae (3%), and WBLF montage captured 96 of 260 vortex ampullae (37%). The superotemporal vortex ampullae was most commonly visualized when compared with the other quadrants (Figure 3) with both camera systems. Only five eyes (8%) from WSLO and no images from the WBLF single image met the UWF consensus definition.
Percentage of visible quadrants by imaging device type. WSLO = widefield confocal scanning laser ophthalmoscope;WBLF = widefield broad line fundus
Quantitative Measurements Between UWF Camera Systems
The average total area captured by the WSLO versus the WBLF single image versus WBLF montage was 781.67 mm2, 433.82 mm2, and 686.03 mm2, respectively (P < .001) (Figure 4). There were statistically significant differences when comparing the average area per individual quadrant. The inferior area was 175.82 on WSLO technology, 102.92 on WBLF single image, and 164.74 on WBLF montage (P ≤ .001), superiorly was 177.70 versus 116.32 versus 177.40, respectively (P < .001), temporally was 217.28 versus 104.48 versus 163.85, respectively (P < .001), and nasally was 222.19 versus 123.24 versus 201.47, respectively (P < .01). When comparing the WSLO versus WBLF montage areas, there were no differences in superior and inferior quadrant (P = .94;P = .06 respectively).
Area of visualized quadrants by imaging device type. WBLF = widefield broad line fundus; WSLO = widefield confocal scanning laser ophthalmoscope
Diagnostic Differences Between UWF Camera Systems
Sixty-five eyes were analyzed on both modalities (WSLO and WBLF montage) regarding clinical grading and diagnoses. There was substantial consensus between graders in both cameras, with an 86% agreement rate for pathology on WSLO images and 88% agreement rate of WBLF images.
Imaging of the peripheral retina has become increasingly relevant for the diagnosis, classification, and management of numerous retinal diseases such as retinovascular diseases such as diabetic retinopathy and retinal vein occlusions. Additional features noticed within the peripheral retina such as areas of ischemia and neovascularization may influence a clinician's treatment approach, thereby highlighting the emerging importance of WF and UWF imaging technology. Although the WSLO and the WBLF cameras are both considered UWF imaging devices, the study results indicate that in routine clinical practice, many images may not meet the consensus definitions of WF or UWF images as proposed by the International Widefield Imaging Group. In our study, only 8% of the WSLO images met the group's definition to be considered as an UWF image, whereas none of the WBLF single-field images met the definition. Furthermore, WSLO images showed the presence of a greater number of vortex ampullae than the WBLF camera both in single-field and montage image formats. This is notable as the consensus group denotes the vortex ampullae as the primary anatomical feature that should be considered when evaluating widefield capabilities.
The quantitative analysis of the images is noteworthy, demonstrating that the total area of the retina image is considerably different between the two instruments. The WSLO covered a significantly larger retina area compared to the WBLF. The WSLO also provided a wider view nasally and temporally. This is consistent with Witmer et al. study, where they analyzed 10 UWF angiography fluorescein images from five patients in two different system that use WSLO technology (Optos Optomap and the Heidelberg Spectralis [Heidelberg Engineering, Heidelberg, Germany]).15 The Optos captured an appreciably wider view of the retina temporally and nasally, albeit with peripheral distortion, than the UWF Heidelberg Spectralis. The larger image capture in the WSLO images likely explains the increased number of visible vortex ampullae visible.
The fact that clinical agreement was similar between both cameras is meaningful, as this indicates that both cameras still provide high-quality images necessary for diagnosing pathology. However, the WSLO system provides a pseudocolor (two-color) image of the retina using the red and green laser wavelengths. The green (red-free) component depicts the retina and its vasculature, whereas the red component highlights deeper structures.10 On the other hand, the WBLF system uses three wide-spectrum LEDs (red, green, and blue channel) to enable image capturing in true color to help enhance the visual contrast of details in certain layers of the retina. In this study, patients were not necessarily chosen based on their potential for having peripheral pathology. Consequently, the additional benefit of being able to examine peripheral regions of the retina is not necessarily well elucidated in this study.
Numerous factors can interfere in the ability to obtain a quality UWF image, such as artifacts created by the adnexa. WSLO was particularly vulnerable to these types of artifacts. This may be related to how the images were acquired (eg, device interface) as well the wider field of view. Additionally, the reduced marginal reflex distance, defined as the vertical reflex between the center of pupil to the margin of superior lid, can also compromise fundus image for both forms of UWF technologies. Generally, these issues were infrequent in this study, but they are noteworthy to consider for ensuring image quality.
The participants of this study were taken in a routine clinical practice, which is valuable because these results represent how these cameras perform outside of a clinical study. Ultimately, images were attempted to be as high-quality as possible by utilizing ophthalmic photographic technicians with significant imaging experience, but images generally seemed to fail to meet International Wide-Field Imaging Study group definition. Consequently, the images still might meet WF and UWF definitions based on field of view measurements, but relatively few might meet a clinically relevant definition that incorporates the identification of useful anatomical features.
Strengths of this study include that its sample of patients routinely seen in a routine retina clinic compared with three different of gauging image differences: area, UWF consensus definitions, and clinical grading. Some of the limitations of the study include its retrospective design, lack of an imaging protocol for all patients, and the fact that investigators could not be truly masked as the images can be differentiated between the various technologies. The authors also acknowledge that a limitation of the analysis is that axial lengths were not recorded and, therefore, a comparison between these cameras in highly myopic eyes is not possible. Additionally, the study does not take into account the warping produced when the retina is projected in a 2-D plane by the different technologies when quantifying the area. In order to produce a WF of view from a single capture, the WSLO camera uses an ellipsoid mirror to image the retina, flattening a three-dimensional curved surface from the retina into a 2-D projection.16,17 As such, the resultant image appears “stretched,” especially on the horizontal axis.18 Recently, however, a stereographic projection software algorithm was introduced to correct for the WSLO peripheral distortion and yield images that maintain the same angular relationship at every eccentricity. The accuracy of the measurements derived from this software has been validated in eyes containing prosthetic implants of known sizes.19 On the other hand, the WBLF uses 133° images and automatically merged to achieve a 200° field of view and produces an undistorted image. Furthermore, this study only tested one specific model of WSLO and WBLF cameras and consequently, results may be variable in other models with similar technologies.
In conclusion, this study examined the utility of a WSLO and WBLF camera by measuring image quadrant areas, the number of visible anatomic features (vortex ampullae), and the pathology elucidated from each image. From a clinical standpoint, it did not seem that there were significant differences between the cameras, but the WSLO camera does show greater potential to identify peripheral features compared to the WBLF.
- Manivannan A, Plskova J, Farrow A, Mckay S, Sharp PF, Forrester JV. Ultra-wide-field fluorescein angiography of the ocular fundus.Am J Ophthalmol. 2005;140(3):525–527. doi:10.1016/j.ajo.2005.02.055PMID:16139004 [CrossRef]
- Spaide RF. Peripheral areas of nonperfusion in treated central retinal vein occlusion as imaged by wide-field fluorescein angiography.Retina. 2011;31(5):829–837. doi:10.1097/IAE.0b013e31820c841ePMID:21487338 [CrossRef]
- Wessel MM, Aaker GD, Parlitsis G, Cho M, D'Amico DJ, Kiss S. Ultra-wide-field angiography improves the detection and classification of diabetic retinopathy.Retina. 2012;32(4):785–791. doi:10.1097/IAE.0b013e3182278b64PMID:22080911 [CrossRef]
- Silva PS, Walia S, Cavallerano JD, et al. Comparison of Low-Light Nonmydriatic Digital Imaging with 35-mm ETDRS Seven-Standard Field Stereo Color Fundus Photographs and Clinical Examination. Telemed J E Health. 2012;18(7):492–499. doi:10.1089/tmj.2011.0232 [CrossRef] PMID: 22827402
- 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(5):949–956. doi:10.1016/j.ophtha.2015.01.008PMID:25704318 [CrossRef]
- Price LD, Au S, Chong NV. Optomap ultrawide field imaging identifies additional retinal abnormalities in patients with diabetic retinopathy.Clin Ophthalmol. 2015;9:527–531. doi:10.2147/OPTH.S79448PMID:25848202 [CrossRef]
- Heussen FM, Tan CS, Sadda SR. Prevalence of peripheral abnormalities on ultra-widefield greenlight (532 nm) autofluorescence imaging at a tertiary care center.Invest Ophthalmol Vis Sci. 2012;53(10):6526–6531. doi:10.1167/iovs.12-9909PMID:22871828 [CrossRef]
- Talks SJ, Manjunath V, Steel DHW, Peto T, Taylor R. New vessels detected on wide-field imaging compared to two-field and seven-field imaging: implications for diabetic retinopathy screening image analysis.Br J Ophthalmol. 2015;99(12):1606–1609. doi:10.1136/bjophthalmol-2015-306719PMID:26271269 [CrossRef]
- Silva PS, Dela Cruz AJ, Ledesma MG, et al. Diabetic Retinopathy Severity and Peripheral Lesions Are Associated with Nonperfusion on Ultrawide Field Angiography.Ophthalmology. 2015;122(12):2465–2472. doi:10.1016/j.ophtha.2015.07.034PMID:26350546 [CrossRef]
- Nagiel A, Lalane RA, Sadda SR, Schwartz SD. ULTRA-WIDEFIELD FUNDUS IMAGING: A Review of Clinical Applications and Future Trends.Retina. 2016;36(4):660–678. doi:10.1097/IAE.0000000000000937PMID:27014860 [CrossRef]
- Mcarthur C. Ultra-wide imaging - True colour ultra-widefield imaging is now a reality in practice.Optician. 2018;(March):1–8.
- CLARUS 500 [package insert]. Dublin, CA: Carl Zeiss Meditec; 2017.
- Choudhry NM. Classification & Guidelines for Wide-field Imaging: Recommendations from the International Wide-field Imaging Study Group. Paper presented at the Retina Society Annual Meeting. ; September 12–15, 2018. ; San Francisco, CA. .
- Escudero-Sanz I, Navarro R. Off-axis aberrations of a wide-angle schematic eye model.J Opt Soc Am A Opt Image Sci Vis. 1999;16(8):1881–1891. doi:10.1364/JOSAA.16.001881PMID:10435267 [CrossRef]
- 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. 2013;7:389–394. doi:10.2147/OPTH.S41731PMID:23458976 [CrossRef]
- Kim JH, Jung H-G, Chung HJ, Lee K, Sohn J. Simplified correction of ischemic index in diabetic retinopathy evaluated by ultra-widefield fluorescein angiography.Korean J Ophthalmol. 2015;29(3):168–172. doi:10.3341/kjo.2015.29.3.168PMID:26028944 [CrossRef]
- Tan CS, Chew MC, van Hemert J, Singer MA, Bell D, Sadda SR. Measuring the precise area of peripheral retinal non-perfusion using ultra-widefield imaging and its correlation with the ischaemic index.Br J Ophthalmol. 2016;100(2):235–239. doi:10.1136/bjophthalmol-2015-306652PMID:26135013 [CrossRef]
- Oishi A, Hidaka J, Yoshimura N. Quantification of the image obtained with a wide-field scanning ophthalmoscope.Invest Ophthalmol Vis Sci. 2014;55(4):2424–2431. doi:10.1167/iovs.13-13738PMID:24667862 [CrossRef]
- Sagong M, van Hemert J, Olmos de Koo LC, Barnett C, Sadda SR. Assessment of accuracy and precision of quantification of ultra-widefield images.Ophthalmology. 2015;122(4):864–866. doi:10.1016/j.ophtha.2014.11.016PMID:25576995 [CrossRef]
|Average age (years)||59|
|Age range (years)||18–84|
|Retinal lattice degeneration||3|
|Age-related macular degeneration||2|
|Other retinal diseases||10|