629158
research-article2016
DSTXXX10.1177/1932296816629158Journal of Diabetes Science and TechnologyMicheletti et al
Special Section
Current and Next Generation Portable Screening Devices for Diabetic Retinopathy
Journal of Diabetes Science and Technology 2016, Vol. 10(2) 295–300 © 2016 Diabetes Technology Society Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1932296816629158 dst.sagepub.com
J. Morgan Micheletti, MD1, Andrew M. Hendrick, MD1, Farah N. Khan, MD2, David C. Ziemer, MD, MPH2, and Francisco J. Pasquel, MD2
Abstract Diabetic retinopathy (DR) is the leading cause of legal blindness in the United States, and with the growing epidemic of diabetes, a global increase in the incidence of DR is inevitable, so it is of utmost importance to identify the most costeffective tools for DR screening. Emerging technology may provide advancements to offset the burden of care, simplify the process, and provide financially responsible methods to safely and effectively optimize care for patients with diabetes mellitus (DM). We review here currently available technology, both in production and under development, for DR screening. Preliminary results of smartphone-based devices, “all-in-one” devices, and alternative technologies are encouraging, but are largely pending verification of utility when used by nonophthalmic personnel. Further research comparing these devices to current nonportable telemedicine strategies and clinical fundus examination is necessary to validate these techniques and to potentially overcome the poor compliance around the globe of current strategies for DR screening. Keywords screening, diabetic retinopathy, smartphone, telemedicine, retina
Diabetic retinopathy (DR) is a common complication of diabetes mellitus (DM) and is a leading cause of blindness worldwide.1 DM is an increasing international health burden that will only increase as the population ages.2,3 Based on individual-level data from population-based studies, the estimated prevalence of any retinopathy among patients with diabetes is around 35% (approximately 93 million people globally).4 Once diagnosed, intensive glycemic and blood pressure control help reduce the risk of DR progression, however achieving recommended targets in DM care has been a major challenge.5 Professional societies recommend either annual or biannual eye exams to screen for DR in patients with DM.6 Compliance with these recommendations is low and the burden outpaces current ability of ophthalmologists to perform screening, especially in geographically distant areas.7 Telemedicine is the most common solution to improve access. It typically involves performance of fundus imaging at point of care with remote interpretation by an eye care provider or trained grader. When applied successfully, systematic screening programs can reduce the risk of blindness. With appropriate implementation, teleretinal screening is both cost-effective and sensitive when compared to ophthalmoscopy.8 Emerging technology may provide advancements to offset the burden of care, simplify the process, and provide
financially responsible methods to safely and effectively optimize care for patients with DM. We review here the currently available technology, both in production and under development, for screening diabetic patients for retinal disease.
Methods We searched and reviewed the literature for current technology in diabetic eye screening that ranged from a minimum production stage of working prototype to full scale production. We contacted the corresponding developers, if necessary, to collect the specifications on each screening device. To focus on the most cost-effective possibilities, devices that cost greater than $10 000, weigh more than 5 pounds, or are 1
Department of Ophthalmology, Emory University School of Medicine, Atlanta, GA, USA 2 Division of Endocrinology, Department of Medicine, Emory University School of Medicine, Atlanta, GA, USA Corresponding Author: Francisco J. Pasquel, MD, Division of Endocrinology, Metabolism & Lipids, Emory University School of Medicine, 49 Jesse Hill Jr Dr SE, Atlanta, GA 30303, USA. Email: [email protected]
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
296
Journal of Diabetes Science and Technology 10(2)
nonportable or dependent on a nonportable device, such as OIS EyeScan, and OptoVue iCam, were excluded. We analyzed and compared each technology’s cost, availability, weight, required equipment, mydriatic versus nonmydriatic capturing, image recording, and availability. After identification of potential devices, data were gathered from the manufacturer, developer, or from their retail product information sheets. The manufacturer’s suggested retail price (MSRP) was gathered either from the company or from available retailers.
Results We identified 9 devices in production or under development, for screening diabetic patients for retinal disease. Comparative studies with standard screening technologies were available only for smartphone-based devices.9,10
Smartphone Fundus Photography Smartphones, available since 1993 when IBM first introduced “the Simon,” have revolutionized communication. A major breakthrough in the smartphone market was the introduction of Apple’s iPhone (Apple Inc, Cupertino, CA, USA) in 2007, followed soon by the introduction of the Android Operating System (Google Inc).11 Smartphones are compelling targets to leverage in teleretinal DR screening due to their ubiquity, portability, low cost, and connectivity for image transfer. With the proper software and limited training, the combination of a smartphone and a 20 diopter lens can capture fundus images using techniques described by Lord et al12 and later improved by Bastawrous13 and Haddock et al.14 An adapter to couple the smartphone with ophthalmoscopy condensing lenses was recently developed by Myung et al.15 The price of this technique, not including the smartphone, is approximately $400 for the lens and the cost of the photography application. Using this method, our group systematically examined 300 subjects (600 eyes) to directly assess sensitivity of DR detection in an Asian Indian cohort.10 Using 7-field photography as the reference standard, we determined that a smartphone under this technique was not highly sensitive for detection of DR.10 This study was performed in a clinical setting to optimize comparisons between photography techniques, not a community-based screening effort. A different smartphone (Samsung Galaxy, Samsung Electronics Co) with an ophthalmoscopy adapter called the Portable Eye Examination Kit (PEEK)16 is being evaluated by Bastawrous et al in collaboration with technologists and public health physicians in low (The Nakuru Eye Disease Cohort Study) and in high income settings.17 This low cost strategy has substantial potential impact in low- and middleincome countries (LMIC) and is currently pending results. The D-Eye device is a magnetically attached portable optical device that attaches to a smartphone (Apple iOS or
Android) and enables visualization of the fundus without the handheld condensing lens. In a comparative study of 120 patients (240 eyes) it was reported to be highly sensitive in detection of DR when a retina specialist viewed the fundus at the point of care through a dilated pupil. The authors of the study propose this technique for teleretinal DR screening purposes.9 The Ocular CellScope is comprised of a mobile phone, a housing that contains the illumination and collections optics, and a holder that aligns the optics with the camera on the phone.18 In a cross-data-set test of a small set of images (N = 30), the authors evaluated a DR screening software that achieved a sensitivity of 100% and a specificity of 80% (AUROC of 0.94 ) for identifying referable DR.19 Optical smartphone extension devices are also available or under development.20 The iExaminer (Welch Allyn), is an FDA-approved device, with an adaptor system that connects a panoptic ophthalmoscope to a smartphone with an integrated iExaminer app to capture images of the retina with a 25-degree field of view (Figure 1). DR screening has not been systematically evaluated with this device.21
Smartphone Applications Smartphone applications are required for some devices (Table 1). An application for server-based telediagnostic analysis for smartphone-based tools is under development, where an automated analysis with a preliminary diagnosis is sent back to the originating smartphone.20 Applications to monitor symptoms are also available, such as myVisionTrack (~$18/month) and DigiSight Sightbook (free application). These 2 smartphone applications allow patients to track small changes to their vision overtime and could potentially serve as a supplement or surrogate to yearly screening exams for diabetic patients. For mass screening, the future of DR screening is heading toward the use of portable devices integrated with automated disease grading software. With PEEK patients could be diagnosed through an automated process. Furthermore smartphone software applications can integrate GPS data that could allow follow up of screened patients.
“All-in-one” Devices The HORUS Scope and SmartScope Pro are devices that provide an all-in-one solution that satisfied inclusion criteria. They have standalone, user-friendly, portable designs that suggest a simplified method to visualize the retina. The HORUS Scope, produced by MiiS (Taiwan), is sold with either a 25-degree lens or a 40-degree lens. While providing a larger field of view, the 40-degree lens is priced at $5200 versus $3100 for the 25-degree lens. The HORUS Scope is handheld, has a built in coaxial light source with infra-red wavelengths for nonmydriatic viewing, and provides highresolution (1080p) video and image capturing. The built in
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
297
Micheletti et al
Figure 1. Current and next generation portable telemedical screening devices for diabetic retinopathy (A) Smartphone and a 20 D lens technique;14 (B) D-Eye Portable Retinal System (www.d-eyecare.com); (C) Portable Eye Examination Kit (PEEK) (www.peekvision.org); (D) SmartScope/Pictor Plus (http://volk.com/pictorplus); (E) Horus Scope (www.jedmed.com); (F) iExaminer (www.welchallyn.com); (G) EyeSelfie;25 (H) Ocular CellScope.18
Li-ion battery allows 3 hours of operating time which could last throughout a day in clinic. The images are stored on a microSD card which allows transfer and access on a PC. A similar device, the SmartScope Pro, branded in the US as Pictor Plus, is an FDA-approved, nonmydriatic, light weight (~1 pound) portable fundus camera (40 degrees field of view) with low flash intensity and capability for video recording. Software with full DICOM and PACS compliance allows exchange across hospital networks.22 With this device, a single fundus image in 56 adult diabetic subjects (n = 111 eyes) was determined to be of sufficient quality to screen for DR by 5 masked ophthalmologists. Compared to dilated clinical examination, the utilization of different grading systems affected the sensitivity and specificity, ranging from 50-85% and 72-98%, respectively.23 A Massachusetts Institute of Technology lab, under direction of Dr Raskar has created 2 compelling technologies that could shape future paradigms in DR screening. One prototype
is called the EyeMitra, which can evaluate both the front and the back of the eye. Composite images up to 120 degrees of peripheral retinal field have been described. No cost or performance data are yet available.24 The other prototype device is called EyeSelfie. This tool appears similar to binoculars and enables users to capture high resolution images of their own retina. The coaxial imaging system uses a display to provides users feedback on optimal alignment. At this time, 10 minutes of training is required to consistently capture well aligned images with a field of view ranging from 15 to 30 degrees. No data yet exist on cost or performance capabilities for this device.25
Alternative Technologies A paradigm shift may be possible to improve screening of DR. The traditional examination of the fundus is optimized with pupillary dilation and flash photography. Furthermore,
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
298
Journal of Diabetes Science and Technology 10(2)
Table 1. Comparison of Portable Telemedical Devices for Diabetic Retinopathy Screening. Device/technology Smartphone and Lens Portable Eye Examination Kit (PEEK) D-Eye Portable Retinal System iExaminer System
Company/developer
Nonmydriatic
Price
Required devices/software Images recorded?
Availability
Multiple
No
$407.99
20 D Lens, Filmic Pro app Yes
Available
Peek collaboration
Yes
~$95
iPhone, Android (pending) Yes
Available
D-Eye
Yes
~$400
Yes
Available
Welch-Allyn
No
$890.99
iPhone 5/5S/6, Samsung Galaxy S4/S5 PanOptic Ophthalmoscope, iExaminer Pro app, iPhone 4/4s Independent Independent
Yes
Available
Yes Yes
Available Available
Horus Scope JEDMED SmartScope Pro/ OptoMed/Volk Optical Pictor Plus Ocular CellScope UC Berkeley
Yes Yes
$4400.00 $9995.00
Yes
N/A
EyeMitra EyeSelfie
Yes Yes
MIT MIT
iPhone 4/4S, DualAlign, i2k Yes Retina software Target $100 Independent N/A Independent
the testing requires skilled individuals in both image acquisition and interpretation. One alternative to circumvent these obstacles is to provide point of care electroretinography (ERG) using a handheld portable device called the RetEval (LKC technologies). ERG abnormalities have been known to precede clinically visible DR26 but widespread use of ERG as a screening tool was not available previously. This was largely due to limited availability—primarily academic medical centers possess the necessary special equipment and expertise to interpret results. The RetEval provides automated interpretation that can determine threshold for referral to ophthalmology for primary care offices. Preliminary reports suggest a 73.7% sensitivity in DR detection, low technical failure rate, and rapid (2.3 minutes average) acquisition time.27,28 Optical coherence tomography (OCT) is the preferred method for evaluating and monitoring diabetic macular edema.29 Handheld OCT technology is currently under development and could potentially become a useful tool for DR screening, particularly macular edema.30
Discussion DR is the most common microvascular complication of DM and is the leading cause of legal blindness in people aged 20-74 years in the United states.31 The onset of DR is estimated to occur about 3 to 7 years before diagnosis of T2D.32,33 Along with the growing epidemic of DM, a global increase in the incidence of DR is inevitable, so it is of utmost importance to identify the most cost-effective tools for DR screening. Telehealth screening modalities for DR have long since been established as crucial strategies in an effort to provide
Development — —
Development Development
cost effective options while also improving patient outcomes.34,35 Even more important than the cost effectiveness of telehealth screening for DR, is the impact that telehealth screening has on overall patient outcomes.36 Nonmydriatic fundus photography was a feasible alternative to direct ophthalmoscopy in a study evaluating both methodologies in an emergency department.37 A study evaluating telehealth retinal screening programs that also incorporated education and care management protocols demonstrated strong correlations between participation in a telehealth eye program and improvement in hemoglobin A1c and LDL levels.38 With the rapid increase in the global prevalence of DM, there is a growing need for telehealth retinal screening devices. To be able to follow current recommendations, screening for DR must be conducted by primary care doctors and diabetes specialists and confirmed by ophthalmologists.6,39 Smartphones will undoubtedly be on the forefront of DR screening in coming years. Technological improvements that enable peripheral visualization of the fundus, standardization of image capture, and improved camera hardware and software will facilitate improved capabilities of smartphone fundus photography. A challenge will be to maintain a continuity with the constant evolution in technology and upgrades in operating systems and devices.21 Current and evolving technologies for DR screening need further investigation. Results from the reviewed technology are encouraging, but are largely pending verification of utility when used by nonophthalmic personnel in primary care settings and of effectiveness compared to current nonportable established telemedicine strategies (eg, 45-degree nonmydriatic photographs), as they are likely operatordependent.9,10,15,21 Tools that integrate smartphones and digital cameras with up-and-coming technologies improve
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
299
Micheletti et al the price point to advance efforts at comprehensive DR screening. The new devices and applications must also ensure that these strategies provide appropriate privacy, confidentiality, and data security measures.40 With constantly progressive development in this cutting edge field, there are sure to be even more advancements for telehealth retinal imaging technologies. Further research is warranted to explore and compare the above devices in a controlled clinical setting to evaluate which devices may provide for the highest sensitivity and effectiveness for DR detection in the hands of ophthalmic versus nonophthalmic personnel. Abbreviations AUROC, area under the receiver operating characteristic curve; DM, diabetes mellitus; DR, diabetic retinopathy; ERG, electroretinography; FDA, Food and Drug Administration; LMIC, low- and middle-income countries; MSRP, manufacturer’s suggested retail price; OCT, optical coherence tomography; PC, personal computer; PEEK, Portable Eye Examination Kit.
Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The author(s) received no financial support for the research, authorship, and/or publication of this article.
References 1. Kobrin Klein BE. Overview of epidemiologic studies of diabetic retinopathy. Ophthal Epidemiol. 2007;14(4):179-183. 2. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010;87(1):4-14. 3. Klein R, Knudtson MD, Lee KE, Gangnon R, Klein BE. The Wisconsin Epidemiologic Study of Diabetic Retinopathy: XXII the twenty-five-year progression of retinopathy in persons with type 1 diabetes. Ophthalmology. 2008;115(11):1859-1868. 4. Yau JW, Rogers SL, Kawasaki R, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012;35:556-564. 5. Haw JS, Tantry S, Vellanki P, Pasquel FJ. National strategies to decrease the burden of diabetes and its complications. Curr Diabetes Rep. 2015;15(9):637. 6. Association AD. 9. Microvascular complications and foot care. Diabetes Care. 2015;38(suppl 1):S58-S66. 7. Bressler NM, Varma R, Doan QV, et al. Underuse of the health care system by persons with diabetes mellitus and diabetic macular edema in the United States. JAMA Ophthalmol. 2014;132(2):168-173. 8. Kirkizlar E, Serban N, Sisson JA, Swann JL, Barnes CS, Williams MD. Evaluation of telemedicine for screening of diabetic retinopathy in the Veterans Health Administration. Ophthalmology. 2013;120(12):2604-2610. 9. Russo A, Morescalchi F, Costagliola C, Delcassi L, Semeraro F. Comparison of smartphone ophthalmoscopy with slit-lamp
biomicroscopy for grading diabetic retinopathy. Am J Ophthalmol. 2015;159(2):360-364. 10. Ryan ME, Rajalakshmi R, Prathiba V, et al. Comparison Among Methods of Retinopathy Assessment (CAMRA) study: smartphone, nonmydriatic, and mydriatic photography. Ophthalmology. 2015;122:2038-2043. 11. Sarwar M, Soomro TR. Impact of smartphones on society. Eur J Sci Res. 2013;98:216-226. 12. Lord RK, Shah VA, San Filippo AN, Krishna R. Novel uses of smartphones in ophthalmology. Ophthalmology. 2010;117(6):1274-1274. 13. Bastawrous A. Smartphone fundoscopy. Ophthalmology. 2012;119(2):432-433. 14. Haddock LJ, Kim DY, Mukai S. Simple, inexpensive technique for high-quality smartphone fundus photography in human and animal eyes. J Ophthalmol. 2013;2013:518479. 15. Myung D, Jais A, He L, Blumenkranz MS, Chang RT. 3D printed smartphone indirect lens adapter for rapid, high quality retinal imaging. J Mobile Tech Med. 2014;3(1):9-15. 16. Portable Eye Examination Kit (PEEK). Available at: http:// www.peekvision.org. Accessed August 12, 2015. 17. Giardini ME, Livingstone IA, Jordan S, et al. A smart phone based ophthalmoscope. IEEE Eng Med Biol Soc. 2014;2014:2177-2180. 18. Maamari RN, Keenan JD, Fletcher DA, Margolis TP. A mobile phone-based retinal camera for portable wide field imaging. Br J Ophthalmol. 2014;98(4):438-441. 19. Bhat S, Bhaskaranand M, Ramachandra C, Margolis TP, Fletcher DA, Solanki K. Fully-automated diabetic retinopathy screening using cellphone-based cameras. Invest Ophthalmol Vis Sci. 2015;56(7):1428-1428. 20. Fink W, Tarbell M. Smart ophthalmics: a smart service platform for tele-ophthalmology. Invest Ophthalmol Vis Sci. 2015;56(7):4110-4110. 21. Pérez GM, Swart W, Munyenyembe JK, Saranchuk P. Barriers to pilot mobile teleophthalmology in a rural hospital in Southern Malawi. Pan African Med J. 2014;19:136. 22. Pictor Plus—digital ophtalmic imager. Available at: http:// volk.com/pictorplus/. Accessed August 12, 2015. 23. Zhang W, Nicholas P, Schuman S, et al. Screening for diabetic retinopathy using the hand-held PICTOR camera. Invest Ophthalmol Vis Sci. 2015;56(7):1426-1426. 24. eyeMITRA: Mobile Retinal Imaging and Predictive Analytics. Available at: http://eyemitra.com. Accessed August 12, 2015. 25. Swedish T, Roesch K, Lee IH, Rastogi K, Bernstein S, Raskar R. EyeSelfie: self directed eye alignment using reciprocal eye box imaging. ACM Trans Graphics. 2015;34(4):58. 26. Shirao Y, Kawasaki K. Electrical responses from diabetic retina. Progress Retinal Eye Res. 1998;17(1):59-76. 27. Kim DB, Ross E, Pham K, et al. New handheld, non-mydriatic ERG device to screen for diabetic retinopathy and other eye diseases. Hemoglobin. 2014;1(7.3):1.3. 28. Maa A, Pillow E, Feuer W, et al. Comparing technical failure rates in diabetic retinopathy screening between RETeval, a 30 Hz flicker electroretinogram device, and mydriatic, 7-field, stereo fundus photography. Diabetes. 2014(suppl 1). Available at: http://app.core-apps.com/tristar-ada14/abstract/ e3048dedfd3d5fbcc516bc86383b513f. 29. Alasil T, Keane PA, Updike JF, et al. Relationship between optical coherence tomography retinal parameters and visual acuity in diabetic macular edema. Ophthalmology. 2010;117(12):2379-2386.
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
300
Journal of Diabetes Science and Technology 10(2)
30. Lu CD, Kraus MF, Potsaid B, et al. Handheld ultrahigh speed swept source optical coherence tomography instrument using a MEMS scanning mirror. Biomed Optics Express. 2014;5(1):293-311. 31. Klein R, Klein BE. Vision disorders in diabetes. Diabetes in America. 1995;1:293. 32. Harris MI, Klein R, Welborn TA, Knuiman MW. Onset of NIDDM occurs at least 4-7 yr before clinical diagnosis. Diabetes Care. 1992;15(7):815-819. 33. Porta M, Curletto G, Cipullo D, et al. Estimating the delay between onset and diagnosis of type 2 diabetes from the time course of retinopathy prevalence. Diabetes Care. 2014;37(6):1668-1674. 34. Li Z, Wu C, Olayiwola JN, Hilaire D, Huang JJ. Telemedicinebased digital retinal imaging vs standard ophthalmologic evaluation for the assessment of diabetic retinopathy. Connecticut Med. 2012;76(2):85-90. 35. Pasquel FJ, Hendrick AM, Ryan M, Cason E, Ali MK, Venkat Narayan KM. Cost-effectiveness of different diabetic retinopathy
screening modalities. J Diabetes Sci Technol. 2016;10(2): 301-307. 36. Bruce BB, Newman NJ, Perez MA, Biousse V. Non-mydriatic ocular fundus photography and telemedicine: past, present, and future. Neuroophthalmology. 2013;37(2). 37. Bruce BB, Lamirel C, Biousse V, et al. Feasibility of nonmydriatic ocular fundus photography in the emergency department: Phase I of the FOTO-ED study. Acad Emerg Med. 2011;18(9):928-933. 38. Fonda SJ, Bursell SE, Lewis DG, Garren J, Hock K, Cavallerano J. The relationship of a diabetes telehealth eye care program to standard eye care and change in diabetes health outcomes. Telemedicine J E-health. 2007;13(6):635-644. 39. MacCuish A. Early detection and screening for diabetic retinopathy. Eye. 1993;7(2):254-259. 40. Kumar S, Wang EH, Pokabla MJ, Noecker RJ. Teleophthalmology assessment of diabetic retinopathy fundus images: smartphone versus standard office computer workstation. Telemedicine J E-health. 2012;18(2):158-162.
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
research-article2016
DSTXXX10.1177/1932296816629158Journal of Diabetes Science and TechnologyMicheletti et al
Special Section
Current and Next Generation Portable Screening Devices for Diabetic Retinopathy
Journal of Diabetes Science and Technology 2016, Vol. 10(2) 295–300 © 2016 Diabetes Technology Society Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1932296816629158 dst.sagepub.com
J. Morgan Micheletti, MD1, Andrew M. Hendrick, MD1, Farah N. Khan, MD2, David C. Ziemer, MD, MPH2, and Francisco J. Pasquel, MD2
Abstract Diabetic retinopathy (DR) is the leading cause of legal blindness in the United States, and with the growing epidemic of diabetes, a global increase in the incidence of DR is inevitable, so it is of utmost importance to identify the most costeffective tools for DR screening. Emerging technology may provide advancements to offset the burden of care, simplify the process, and provide financially responsible methods to safely and effectively optimize care for patients with diabetes mellitus (DM). We review here currently available technology, both in production and under development, for DR screening. Preliminary results of smartphone-based devices, “all-in-one” devices, and alternative technologies are encouraging, but are largely pending verification of utility when used by nonophthalmic personnel. Further research comparing these devices to current nonportable telemedicine strategies and clinical fundus examination is necessary to validate these techniques and to potentially overcome the poor compliance around the globe of current strategies for DR screening. Keywords screening, diabetic retinopathy, smartphone, telemedicine, retina
Diabetic retinopathy (DR) is a common complication of diabetes mellitus (DM) and is a leading cause of blindness worldwide.1 DM is an increasing international health burden that will only increase as the population ages.2,3 Based on individual-level data from population-based studies, the estimated prevalence of any retinopathy among patients with diabetes is around 35% (approximately 93 million people globally).4 Once diagnosed, intensive glycemic and blood pressure control help reduce the risk of DR progression, however achieving recommended targets in DM care has been a major challenge.5 Professional societies recommend either annual or biannual eye exams to screen for DR in patients with DM.6 Compliance with these recommendations is low and the burden outpaces current ability of ophthalmologists to perform screening, especially in geographically distant areas.7 Telemedicine is the most common solution to improve access. It typically involves performance of fundus imaging at point of care with remote interpretation by an eye care provider or trained grader. When applied successfully, systematic screening programs can reduce the risk of blindness. With appropriate implementation, teleretinal screening is both cost-effective and sensitive when compared to ophthalmoscopy.8 Emerging technology may provide advancements to offset the burden of care, simplify the process, and provide
financially responsible methods to safely and effectively optimize care for patients with DM. We review here the currently available technology, both in production and under development, for screening diabetic patients for retinal disease.
Methods We searched and reviewed the literature for current technology in diabetic eye screening that ranged from a minimum production stage of working prototype to full scale production. We contacted the corresponding developers, if necessary, to collect the specifications on each screening device. To focus on the most cost-effective possibilities, devices that cost greater than $10 000, weigh more than 5 pounds, or are 1
Department of Ophthalmology, Emory University School of Medicine, Atlanta, GA, USA 2 Division of Endocrinology, Department of Medicine, Emory University School of Medicine, Atlanta, GA, USA Corresponding Author: Francisco J. Pasquel, MD, Division of Endocrinology, Metabolism & Lipids, Emory University School of Medicine, 49 Jesse Hill Jr Dr SE, Atlanta, GA 30303, USA. Email: [email protected]
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
296
Journal of Diabetes Science and Technology 10(2)
nonportable or dependent on a nonportable device, such as OIS EyeScan, and OptoVue iCam, were excluded. We analyzed and compared each technology’s cost, availability, weight, required equipment, mydriatic versus nonmydriatic capturing, image recording, and availability. After identification of potential devices, data were gathered from the manufacturer, developer, or from their retail product information sheets. The manufacturer’s suggested retail price (MSRP) was gathered either from the company or from available retailers.
Results We identified 9 devices in production or under development, for screening diabetic patients for retinal disease. Comparative studies with standard screening technologies were available only for smartphone-based devices.9,10
Smartphone Fundus Photography Smartphones, available since 1993 when IBM first introduced “the Simon,” have revolutionized communication. A major breakthrough in the smartphone market was the introduction of Apple’s iPhone (Apple Inc, Cupertino, CA, USA) in 2007, followed soon by the introduction of the Android Operating System (Google Inc).11 Smartphones are compelling targets to leverage in teleretinal DR screening due to their ubiquity, portability, low cost, and connectivity for image transfer. With the proper software and limited training, the combination of a smartphone and a 20 diopter lens can capture fundus images using techniques described by Lord et al12 and later improved by Bastawrous13 and Haddock et al.14 An adapter to couple the smartphone with ophthalmoscopy condensing lenses was recently developed by Myung et al.15 The price of this technique, not including the smartphone, is approximately $400 for the lens and the cost of the photography application. Using this method, our group systematically examined 300 subjects (600 eyes) to directly assess sensitivity of DR detection in an Asian Indian cohort.10 Using 7-field photography as the reference standard, we determined that a smartphone under this technique was not highly sensitive for detection of DR.10 This study was performed in a clinical setting to optimize comparisons between photography techniques, not a community-based screening effort. A different smartphone (Samsung Galaxy, Samsung Electronics Co) with an ophthalmoscopy adapter called the Portable Eye Examination Kit (PEEK)16 is being evaluated by Bastawrous et al in collaboration with technologists and public health physicians in low (The Nakuru Eye Disease Cohort Study) and in high income settings.17 This low cost strategy has substantial potential impact in low- and middleincome countries (LMIC) and is currently pending results. The D-Eye device is a magnetically attached portable optical device that attaches to a smartphone (Apple iOS or
Android) and enables visualization of the fundus without the handheld condensing lens. In a comparative study of 120 patients (240 eyes) it was reported to be highly sensitive in detection of DR when a retina specialist viewed the fundus at the point of care through a dilated pupil. The authors of the study propose this technique for teleretinal DR screening purposes.9 The Ocular CellScope is comprised of a mobile phone, a housing that contains the illumination and collections optics, and a holder that aligns the optics with the camera on the phone.18 In a cross-data-set test of a small set of images (N = 30), the authors evaluated a DR screening software that achieved a sensitivity of 100% and a specificity of 80% (AUROC of 0.94 ) for identifying referable DR.19 Optical smartphone extension devices are also available or under development.20 The iExaminer (Welch Allyn), is an FDA-approved device, with an adaptor system that connects a panoptic ophthalmoscope to a smartphone with an integrated iExaminer app to capture images of the retina with a 25-degree field of view (Figure 1). DR screening has not been systematically evaluated with this device.21
Smartphone Applications Smartphone applications are required for some devices (Table 1). An application for server-based telediagnostic analysis for smartphone-based tools is under development, where an automated analysis with a preliminary diagnosis is sent back to the originating smartphone.20 Applications to monitor symptoms are also available, such as myVisionTrack (~$18/month) and DigiSight Sightbook (free application). These 2 smartphone applications allow patients to track small changes to their vision overtime and could potentially serve as a supplement or surrogate to yearly screening exams for diabetic patients. For mass screening, the future of DR screening is heading toward the use of portable devices integrated with automated disease grading software. With PEEK patients could be diagnosed through an automated process. Furthermore smartphone software applications can integrate GPS data that could allow follow up of screened patients.
“All-in-one” Devices The HORUS Scope and SmartScope Pro are devices that provide an all-in-one solution that satisfied inclusion criteria. They have standalone, user-friendly, portable designs that suggest a simplified method to visualize the retina. The HORUS Scope, produced by MiiS (Taiwan), is sold with either a 25-degree lens or a 40-degree lens. While providing a larger field of view, the 40-degree lens is priced at $5200 versus $3100 for the 25-degree lens. The HORUS Scope is handheld, has a built in coaxial light source with infra-red wavelengths for nonmydriatic viewing, and provides highresolution (1080p) video and image capturing. The built in
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
297
Micheletti et al
Figure 1. Current and next generation portable telemedical screening devices for diabetic retinopathy (A) Smartphone and a 20 D lens technique;14 (B) D-Eye Portable Retinal System (www.d-eyecare.com); (C) Portable Eye Examination Kit (PEEK) (www.peekvision.org); (D) SmartScope/Pictor Plus (http://volk.com/pictorplus); (E) Horus Scope (www.jedmed.com); (F) iExaminer (www.welchallyn.com); (G) EyeSelfie;25 (H) Ocular CellScope.18
Li-ion battery allows 3 hours of operating time which could last throughout a day in clinic. The images are stored on a microSD card which allows transfer and access on a PC. A similar device, the SmartScope Pro, branded in the US as Pictor Plus, is an FDA-approved, nonmydriatic, light weight (~1 pound) portable fundus camera (40 degrees field of view) with low flash intensity and capability for video recording. Software with full DICOM and PACS compliance allows exchange across hospital networks.22 With this device, a single fundus image in 56 adult diabetic subjects (n = 111 eyes) was determined to be of sufficient quality to screen for DR by 5 masked ophthalmologists. Compared to dilated clinical examination, the utilization of different grading systems affected the sensitivity and specificity, ranging from 50-85% and 72-98%, respectively.23 A Massachusetts Institute of Technology lab, under direction of Dr Raskar has created 2 compelling technologies that could shape future paradigms in DR screening. One prototype
is called the EyeMitra, which can evaluate both the front and the back of the eye. Composite images up to 120 degrees of peripheral retinal field have been described. No cost or performance data are yet available.24 The other prototype device is called EyeSelfie. This tool appears similar to binoculars and enables users to capture high resolution images of their own retina. The coaxial imaging system uses a display to provides users feedback on optimal alignment. At this time, 10 minutes of training is required to consistently capture well aligned images with a field of view ranging from 15 to 30 degrees. No data yet exist on cost or performance capabilities for this device.25
Alternative Technologies A paradigm shift may be possible to improve screening of DR. The traditional examination of the fundus is optimized with pupillary dilation and flash photography. Furthermore,
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
298
Journal of Diabetes Science and Technology 10(2)
Table 1. Comparison of Portable Telemedical Devices for Diabetic Retinopathy Screening. Device/technology Smartphone and Lens Portable Eye Examination Kit (PEEK) D-Eye Portable Retinal System iExaminer System
Company/developer
Nonmydriatic
Price
Required devices/software Images recorded?
Availability
Multiple
No
$407.99
20 D Lens, Filmic Pro app Yes
Available
Peek collaboration
Yes
~$95
iPhone, Android (pending) Yes
Available
D-Eye
Yes
~$400
Yes
Available
Welch-Allyn
No
$890.99
iPhone 5/5S/6, Samsung Galaxy S4/S5 PanOptic Ophthalmoscope, iExaminer Pro app, iPhone 4/4s Independent Independent
Yes
Available
Yes Yes
Available Available
Horus Scope JEDMED SmartScope Pro/ OptoMed/Volk Optical Pictor Plus Ocular CellScope UC Berkeley
Yes Yes
$4400.00 $9995.00
Yes
N/A
EyeMitra EyeSelfie
Yes Yes
MIT MIT
iPhone 4/4S, DualAlign, i2k Yes Retina software Target $100 Independent N/A Independent
the testing requires skilled individuals in both image acquisition and interpretation. One alternative to circumvent these obstacles is to provide point of care electroretinography (ERG) using a handheld portable device called the RetEval (LKC technologies). ERG abnormalities have been known to precede clinically visible DR26 but widespread use of ERG as a screening tool was not available previously. This was largely due to limited availability—primarily academic medical centers possess the necessary special equipment and expertise to interpret results. The RetEval provides automated interpretation that can determine threshold for referral to ophthalmology for primary care offices. Preliminary reports suggest a 73.7% sensitivity in DR detection, low technical failure rate, and rapid (2.3 minutes average) acquisition time.27,28 Optical coherence tomography (OCT) is the preferred method for evaluating and monitoring diabetic macular edema.29 Handheld OCT technology is currently under development and could potentially become a useful tool for DR screening, particularly macular edema.30
Discussion DR is the most common microvascular complication of DM and is the leading cause of legal blindness in people aged 20-74 years in the United states.31 The onset of DR is estimated to occur about 3 to 7 years before diagnosis of T2D.32,33 Along with the growing epidemic of DM, a global increase in the incidence of DR is inevitable, so it is of utmost importance to identify the most cost-effective tools for DR screening. Telehealth screening modalities for DR have long since been established as crucial strategies in an effort to provide
Development — —
Development Development
cost effective options while also improving patient outcomes.34,35 Even more important than the cost effectiveness of telehealth screening for DR, is the impact that telehealth screening has on overall patient outcomes.36 Nonmydriatic fundus photography was a feasible alternative to direct ophthalmoscopy in a study evaluating both methodologies in an emergency department.37 A study evaluating telehealth retinal screening programs that also incorporated education and care management protocols demonstrated strong correlations between participation in a telehealth eye program and improvement in hemoglobin A1c and LDL levels.38 With the rapid increase in the global prevalence of DM, there is a growing need for telehealth retinal screening devices. To be able to follow current recommendations, screening for DR must be conducted by primary care doctors and diabetes specialists and confirmed by ophthalmologists.6,39 Smartphones will undoubtedly be on the forefront of DR screening in coming years. Technological improvements that enable peripheral visualization of the fundus, standardization of image capture, and improved camera hardware and software will facilitate improved capabilities of smartphone fundus photography. A challenge will be to maintain a continuity with the constant evolution in technology and upgrades in operating systems and devices.21 Current and evolving technologies for DR screening need further investigation. Results from the reviewed technology are encouraging, but are largely pending verification of utility when used by nonophthalmic personnel in primary care settings and of effectiveness compared to current nonportable established telemedicine strategies (eg, 45-degree nonmydriatic photographs), as they are likely operatordependent.9,10,15,21 Tools that integrate smartphones and digital cameras with up-and-coming technologies improve
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
299
Micheletti et al the price point to advance efforts at comprehensive DR screening. The new devices and applications must also ensure that these strategies provide appropriate privacy, confidentiality, and data security measures.40 With constantly progressive development in this cutting edge field, there are sure to be even more advancements for telehealth retinal imaging technologies. Further research is warranted to explore and compare the above devices in a controlled clinical setting to evaluate which devices may provide for the highest sensitivity and effectiveness for DR detection in the hands of ophthalmic versus nonophthalmic personnel. Abbreviations AUROC, area under the receiver operating characteristic curve; DM, diabetes mellitus; DR, diabetic retinopathy; ERG, electroretinography; FDA, Food and Drug Administration; LMIC, low- and middle-income countries; MSRP, manufacturer’s suggested retail price; OCT, optical coherence tomography; PC, personal computer; PEEK, Portable Eye Examination Kit.
Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The author(s) received no financial support for the research, authorship, and/or publication of this article.
References 1. Kobrin Klein BE. Overview of epidemiologic studies of diabetic retinopathy. Ophthal Epidemiol. 2007;14(4):179-183. 2. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010;87(1):4-14. 3. Klein R, Knudtson MD, Lee KE, Gangnon R, Klein BE. The Wisconsin Epidemiologic Study of Diabetic Retinopathy: XXII the twenty-five-year progression of retinopathy in persons with type 1 diabetes. Ophthalmology. 2008;115(11):1859-1868. 4. Yau JW, Rogers SL, Kawasaki R, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012;35:556-564. 5. Haw JS, Tantry S, Vellanki P, Pasquel FJ. National strategies to decrease the burden of diabetes and its complications. Curr Diabetes Rep. 2015;15(9):637. 6. Association AD. 9. Microvascular complications and foot care. Diabetes Care. 2015;38(suppl 1):S58-S66. 7. Bressler NM, Varma R, Doan QV, et al. Underuse of the health care system by persons with diabetes mellitus and diabetic macular edema in the United States. JAMA Ophthalmol. 2014;132(2):168-173. 8. Kirkizlar E, Serban N, Sisson JA, Swann JL, Barnes CS, Williams MD. Evaluation of telemedicine for screening of diabetic retinopathy in the Veterans Health Administration. Ophthalmology. 2013;120(12):2604-2610. 9. Russo A, Morescalchi F, Costagliola C, Delcassi L, Semeraro F. Comparison of smartphone ophthalmoscopy with slit-lamp
biomicroscopy for grading diabetic retinopathy. Am J Ophthalmol. 2015;159(2):360-364. 10. Ryan ME, Rajalakshmi R, Prathiba V, et al. Comparison Among Methods of Retinopathy Assessment (CAMRA) study: smartphone, nonmydriatic, and mydriatic photography. Ophthalmology. 2015;122:2038-2043. 11. Sarwar M, Soomro TR. Impact of smartphones on society. Eur J Sci Res. 2013;98:216-226. 12. Lord RK, Shah VA, San Filippo AN, Krishna R. Novel uses of smartphones in ophthalmology. Ophthalmology. 2010;117(6):1274-1274. 13. Bastawrous A. Smartphone fundoscopy. Ophthalmology. 2012;119(2):432-433. 14. Haddock LJ, Kim DY, Mukai S. Simple, inexpensive technique for high-quality smartphone fundus photography in human and animal eyes. J Ophthalmol. 2013;2013:518479. 15. Myung D, Jais A, He L, Blumenkranz MS, Chang RT. 3D printed smartphone indirect lens adapter for rapid, high quality retinal imaging. J Mobile Tech Med. 2014;3(1):9-15. 16. Portable Eye Examination Kit (PEEK). Available at: http:// www.peekvision.org. Accessed August 12, 2015. 17. Giardini ME, Livingstone IA, Jordan S, et al. A smart phone based ophthalmoscope. IEEE Eng Med Biol Soc. 2014;2014:2177-2180. 18. Maamari RN, Keenan JD, Fletcher DA, Margolis TP. A mobile phone-based retinal camera for portable wide field imaging. Br J Ophthalmol. 2014;98(4):438-441. 19. Bhat S, Bhaskaranand M, Ramachandra C, Margolis TP, Fletcher DA, Solanki K. Fully-automated diabetic retinopathy screening using cellphone-based cameras. Invest Ophthalmol Vis Sci. 2015;56(7):1428-1428. 20. Fink W, Tarbell M. Smart ophthalmics: a smart service platform for tele-ophthalmology. Invest Ophthalmol Vis Sci. 2015;56(7):4110-4110. 21. Pérez GM, Swart W, Munyenyembe JK, Saranchuk P. Barriers to pilot mobile teleophthalmology in a rural hospital in Southern Malawi. Pan African Med J. 2014;19:136. 22. Pictor Plus—digital ophtalmic imager. Available at: http:// volk.com/pictorplus/. Accessed August 12, 2015. 23. Zhang W, Nicholas P, Schuman S, et al. Screening for diabetic retinopathy using the hand-held PICTOR camera. Invest Ophthalmol Vis Sci. 2015;56(7):1426-1426. 24. eyeMITRA: Mobile Retinal Imaging and Predictive Analytics. Available at: http://eyemitra.com. Accessed August 12, 2015. 25. Swedish T, Roesch K, Lee IH, Rastogi K, Bernstein S, Raskar R. EyeSelfie: self directed eye alignment using reciprocal eye box imaging. ACM Trans Graphics. 2015;34(4):58. 26. Shirao Y, Kawasaki K. Electrical responses from diabetic retina. Progress Retinal Eye Res. 1998;17(1):59-76. 27. Kim DB, Ross E, Pham K, et al. New handheld, non-mydriatic ERG device to screen for diabetic retinopathy and other eye diseases. Hemoglobin. 2014;1(7.3):1.3. 28. Maa A, Pillow E, Feuer W, et al. Comparing technical failure rates in diabetic retinopathy screening between RETeval, a 30 Hz flicker electroretinogram device, and mydriatic, 7-field, stereo fundus photography. Diabetes. 2014(suppl 1). Available at: http://app.core-apps.com/tristar-ada14/abstract/ e3048dedfd3d5fbcc516bc86383b513f. 29. Alasil T, Keane PA, Updike JF, et al. Relationship between optical coherence tomography retinal parameters and visual acuity in diabetic macular edema. Ophthalmology. 2010;117(12):2379-2386.
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
300
Journal of Diabetes Science and Technology 10(2)
30. Lu CD, Kraus MF, Potsaid B, et al. Handheld ultrahigh speed swept source optical coherence tomography instrument using a MEMS scanning mirror. Biomed Optics Express. 2014;5(1):293-311. 31. Klein R, Klein BE. Vision disorders in diabetes. Diabetes in America. 1995;1:293. 32. Harris MI, Klein R, Welborn TA, Knuiman MW. Onset of NIDDM occurs at least 4-7 yr before clinical diagnosis. Diabetes Care. 1992;15(7):815-819. 33. Porta M, Curletto G, Cipullo D, et al. Estimating the delay between onset and diagnosis of type 2 diabetes from the time course of retinopathy prevalence. Diabetes Care. 2014;37(6):1668-1674. 34. Li Z, Wu C, Olayiwola JN, Hilaire D, Huang JJ. Telemedicinebased digital retinal imaging vs standard ophthalmologic evaluation for the assessment of diabetic retinopathy. Connecticut Med. 2012;76(2):85-90. 35. Pasquel FJ, Hendrick AM, Ryan M, Cason E, Ali MK, Venkat Narayan KM. Cost-effectiveness of different diabetic retinopathy
screening modalities. J Diabetes Sci Technol. 2016;10(2): 301-307. 36. Bruce BB, Newman NJ, Perez MA, Biousse V. Non-mydriatic ocular fundus photography and telemedicine: past, present, and future. Neuroophthalmology. 2013;37(2). 37. Bruce BB, Lamirel C, Biousse V, et al. Feasibility of nonmydriatic ocular fundus photography in the emergency department: Phase I of the FOTO-ED study. Acad Emerg Med. 2011;18(9):928-933. 38. Fonda SJ, Bursell SE, Lewis DG, Garren J, Hock K, Cavallerano J. The relationship of a diabetes telehealth eye care program to standard eye care and change in diabetes health outcomes. Telemedicine J E-health. 2007;13(6):635-644. 39. MacCuish A. Early detection and screening for diabetic retinopathy. Eye. 1993;7(2):254-259. 40. Kumar S, Wang EH, Pokabla MJ, Noecker RJ. Teleophthalmology assessment of diabetic retinopathy fundus images: smartphone versus standard office computer workstation. Telemedicine J E-health. 2012;18(2):158-162.
Downloaded from dst.sagepub.com at Bobst Library, New York University on March 6, 2016
