Advertisement

PentaVision

Retinal Physician

Article

Update on Applications of Teleophthalmology in Retina

Home monitoring and AI promise to bring ophthalmology forward in telemedicine.

By: Dinah Chen, MD, Yasha Modi, MD , Lama A. Al-Aswad, MD, MPH
Retinal Physician November 1, 2021 Vol 18, Issue November/December 2021 Page(s): 47, 48, 50, 51

Related

Telemedicine has increasingly become a well-accepted form of health-care delivery in the United States, across all medical specialties. At the height of the pandemic in 2020, telemedicine visits increased to 48% of total visits in some states.1 The trend in digital health initiatives continues to strengthen not only across academic institutions but also in the private sector. Venture funding for digital health in Q1 2021 reached an unprecedented high of $7.2 billion, and it is expected to continue its impressive growth, particularly among telemedicine initiatives and mobile health apps.2

Although teleophthalmology is not a new concept, it is only in the last year that synchronous, video, and telephone-based eye visits have risen significantly.3 The pandemic has spurred shifting attitudes toward digital transformation in ophthalmology, across subspecialities in large part due to expanded reimbursement for televisits.4 An international survey of more than 200 retina specialists found an increase in teleophthalmology during this period with 42.1% of institutions reporting increased home monitoring of hyperacuity, metamorphopsias, and scotomas specifically.5

Not only does teleophthalmology have the potential to expand access to care and bring ophthalmic care to regions with physician shortages or poor coverage, it also represents an opportunity to reduce health-care costs through scalable screening of vision-threatening conditions.6 With increased capacity for remote imaging and testing, and integration of artificial intelligence (AI), synchronous retinal telemedicine is quickly becoming an emerging possibility.

STORE-AND-FORWARD TELEOPHTHALMOLOGY

Historically, teleophthalmology and specifically, teleretina has taken the form of store-and-forward, or asynchronous evaluation. Under this model, diagnostic imaging and other testing is performed, often in primary care clinics, and sent to an outside reader. There is no associated in-person visit with an eye provider. One such robust and well-established design is the TECS program, developed at the Atlanta VA Medical Center. This program utilizes trained ophthalmic technicians to perform vision checks and assist in diagnostic imaging screening for vision-threatening diseases with a focus on cataract, glaucoma, macular degeneration, and diabetic retinopathy (DR).7 Approximately 30% of patients screened through the TECS program are referred to the eye clinic for additional evaluation, 10% of this population for retinal findings.8 Many institutions have developed similar store-and-forward programs throughout the country, particularly for the screening of diabetic eye disease. The University of Wisconsin–Madison has an established teleophthalmology program through its primary care clinics. Similarly, Duke University, New York University, and Stanford have telescreening specifically for diabetic referrals.

Store-and-forward methodology has also been employed successfully for retinopathy of prematurity (ROP) screening. Stanford’s well-established initiative, SUNDROP (Stanford University Network for Diagnosis of Retinopathy of Prematurity), was started in 2005 and has demonstrated high diagnostic accuracy — 100% sensitivity and 99.8% specificity — of treatment-warranted ROP using telemedicine.9 This teleophthalmology protocol has also been successfully used to follow infants receiving laser therapy for ROP.10

Although asynchronous visits for retinal pathologies will likely continue to represent a significant portion of teleophthalmology in the future, advances in AI are already beginning to change the landscape of digital retinal evaluation and treatment.

ARTIFICIAL INTELLIGENCE FOR RETINAL DISEASE SCREENING

Artificial intelligence represents an emerging adjunct to telemedicine and is quickly becoming clinically relevant in ophthalmology. In 2018, IDx-DR became the first autonomous AI to be approved by the FDA, for the detection of referable DR.11 This was followed by Eyenuk’s FDA approval for an autonomous AI identifying more-than-mild and vision threatening DR in 2020.12 These devices have ushered in a new era of retinal screening, removing some of the most time-consuming steps from the traditional store-and-forward model. This technology is autonomous, meaning it does not necessitate image interpretation by an eye provider and both devices reported high imageability as well as high sensitivity and specificity for identifying DR.12,13 In IDx-DR’s clinical trial for FDA approval, roughly 24% of patients were found to have referrable DR, 198 out of the 819 patients screened.13 In the United States, there are 34.2 million Americans with diabetes, and, as rates continue to rise, screening measures powered by AI will be essential to meeting this need.

Both IDx-Dr and EyeArt have marketed their algorithm software tied to physical devices, the Topcon and Canon fundus cameras, respectively. Although they remain the only FDA-approved autonomous AI in ophthalmology, numerous other AI algorithms have been developed for fundus photo, fundus autofluorescence, and OCT-based diagnosis of various retinal conditions including DR, diabetic macular edema (DME), age-related macular degeneration (AMD), ROP, and retinitis pigmentosa (RP).14,15 Many of these algorithms are highly accurate; in fact, in 2018, Brown et al published research showing that a deep convolutional neural network was able to outperform ROP experts in the diagnosis of plus disease.16

Beyond diagnosis alone, research in AI is being conducted to identify lesion activity and possibly guide treat and extend protocols for intravitreal VEGF therapies for both AMD and DME. In a recent study, Mantel et al developed a highly sensitive and specific deep learning algorithm for the detection of intraretinal fluid, subretinal fluid, and pigment epithelial detachments in patients with wet AMD.17 Prahs et al developed a deep learning algorithm that accurately predicted whether a patient received an intravitreal injection based on OCT images alone for various conditions.18 Artificial intelligence, in combination with novel ways of imaging the retina, has the potential to shape both where and how we diagnose and follow patients. Much attention is now being turned toward the application of AI to images obtained on mobile devices.

MOBILE DEVICE IMAGING

Several smartphone adapters for fundus photography are commercially available, and many groups have published research on low-cost prototypes for homemade adapters. Toslak et al designed a low-cost 3D-printed smartphone adapter capable of taking widefield fundus images.19 Studies have demonstrated these smart phone images to be reliable for the detection of DR.20 Others have studied the use of smartphone images in diagnosis of pediatric retinal conditions; Patel et al reported good agreement between diagnoses from photos taken with a widefield smartphone adapter and treating clinicians for diseases such as retinoblastoma, Coats disease, and commotio retinae.21

Artificial intelligence for the detection of DR has been reliably applied to fundus images acquired on smartphones.22,23 In a study utilizing a commercially available smartphone lens adapter, fundus videos were taken by persons at various levels of medical training and low-resolution screenshots were evaluated with an AI that was able to correctly identify moderate or worse DR.24 Smartphone-based imaging with integrated AI represents a scalable solution for screening, especially in resource-limited settings.

HOME MONITORING

Although other fields in medicine have well-established home-monitoring as adjuncts to their telemedicine efforts, such as continuous glucose monitoring, home lung function testing, and blood sampling, ophthalmology has yet to solidify remote monitoring as a part of disease management. In the last decade, new devices have come to market that are suitable for home use, including devices for AMD and DME monitoring in particular.

Mobile Apps

This year, the ophthalmology innovation lab at NYU led by Dr. Al-Aswad launched its Eye Test app (NYU Langone Eye Test), which assesses near visual acuity with full integration into the EPIC electronic medical record (EMR). It is currently undergoing validation studies, but early results demonstrate it to be comparable to Rosenbaum visual acuity testing. Other mobile applications are also available for visual acuity, Amsler grid, and other tests, but these need further validation and are yet to be integrated with the EMR.

Alleye (Oculocare Medical) and myVisiontrack (Genentech) are 2 FDA-approved, mobile-phone-based apps for hyperacuity monitoring that have demonstrated effectiveness in early detection and progression of both AMD and DME. In a recent study, patients undergoing intravitreal injection therapy for wet AMD or DME were able to self-identify disease progression at home using Alleye.25 Clinicians were alerted to changes in patient scores, and if persistent changes in vision were noted, patient follow-up appointments were expedited. The app achieved a positive predictive value of 80% when compared with anatomical progression at follow-up visits.25 Notal Vision’s ForeseeHome is an FDA-approved device that utilizes hyperacuity and macular visual field testing to detect choroidal neovascularization. The randomized trial for this device was terminated early given positive findings; the device arm of high-risk AMD patients lost fewer letters than those in the standard arm.26

Home-Based Devices

Home-based optical coherence tomography (OCT) devices are increasingly becoming a reality and are likely to broaden the scope of home monitoring. Notal Vision has developed a home-based OCT device with integrated AI for the detection of wet AMD that is currently undergoing clinical trial testing.27 Visotec is a German company developing hand-held OCT machines. In 2020, after testing the prototype, Visotec reported a sensitivity and specificity of 94% and 95%, respectively, for anti-VEGF treatment decisions.28 They anticipate that by 2024, the product will be commercially available.29 Other groups have also developed portable OCT devices; in 2019 researchers at Duke University presented data on their low-cost 3D-printed OCT, which demonstrated a contrast-to-noise signal similar to standard OCTs.30 These promising devices could reshape not only diagnosis but also follow-up care for retina patients. In general, as medicine moves toward increased transparency for patients, home monitoring will hopefully improve quality of care and empower patients to understand their own diseases processes.

FUTURE TRENDS

The rapid growth in telemedicine in the private sector signals growing demand among consumers for the digitization of health care. Ophthalmology has often been at the forefront of technological innovation but it remains one of the fields within medicine with the lowest rates of telemedicine use, even despite major growth during the pandemic.1,31,32 At one institution, ophthalmology televisits accounted for less than 1% total ophthalmology visits while approximately 75% of neurosurgery visits were virtual.32 Historically, the cost and accessibility of ophthalmologic equipment and poor reimbursement rates have been listed as significant barriers to adoption.33 In the future, increased ability to monitor patients at home coupled with dramatic growth in AI promise to improve efficiency in the field of ophthalmology and retina specifically, facilitating the movement toward a truly synchronous model of telemedicine. RP

Dinah Chen, MD, is an innovation fellow and clinical instructor of ophthalmology at NYU Langone Health in New York, New York. She reports consultancy to Clover Therapeutics. Yasha Modi, MD, is an associate professor and director of teleretina in the department of ophthalmology at NYU Grossman School of Medicine in New York, New York. Dr. Modi reports consultancy to Allergan, Alimera, Genentech, Thea, and Zeiss. Lama A. Al-Aswad, MD, MPH, is the director of the ophthalmology innovation fellowship at NYU Langone Health. Dr. Al-Aswad reports grant or research support from New World Medical and Save Vision Foundation; consultancy or expenses from Zeiss and Topcon; stock ownership and advisory board membership with Save Vision Foundation, and other financial interests with GlobeChek and AI Optics. Reach Dr. Al-Aswad at lama.al-aswad@nyulangone.org.

REFERENCES

  1. Patel SY, Mehrotra A, Huskamp HA, Uscher-Pines L, Ganguli I, Barnett ML. Trends in outpatient care delivery and telemedicine during the COVID-19 pandemic in the US. JAMA Intern Med. 2021;181(3):388-391. doi:10.1001/jamainternmed.2020.5928
  2. Balasubramanian S. An ode to digital health: The US government is investing $80 million to create a new public health informatics & technology program. Forbes. 2021. Accessed September 28, 2021. https://www.forbes.com/sites/saibala/2021/06/21/an-ode-to-digital-health-the-us-government-is-investing-80-million-to-create-a-new-public-health-informatics--technology-program
  3. Portney DS, Zhu Z, Chen EM, et al. COVID-19 and use of teleophthalmology (CUT group): trends and diagnoses. Ophthalmology. 2021;128(10):1483-1485. doi:10.1016/j.ophtha.2021.02.010
  4. De Lott LB, Newman-Casey PA, Lee PP, et al. Change in ophthalmic clinicians’ attitudes toward telemedicine during the coronavirus 2019 pandemic. Telemed J E Health. 2021;27(2):231-235. doi:10.1089/tmj.2020.0222
  5. Faes L, Rosenblatt A, Schwartz R, et al. Overcoming barriers of retinal care delivery during a pandemic-attitudes and drivers for the implementation of digital health: a global expert survey [published online ahead of print, 2020 Oct 16]. Br J Ophthalmol. 2020;bjophthalmol-2020-316882. doi:10.1136/bjophthalmol-2020-316882
  6. Sharafeldin N, Kawaguchi A, Sundaram A, et al. Review of economic evaluations of teleophthalmology as a screening strategy for chronic eye disease in adults. Br J Ophthalmol. 2018;102(11):1485-1491. doi:10.1136/bjophthalmol-2017-311452
  7. American Academy of Ophthalmology. Technology-based eye care services in the Department of Veterans Affairs. Accessed September 28, 2021. https://www.aao.org/Assets/2f74bebb-f24d-4eaf-a9ac-3913e3b8ac5f/636589766568530000/va-tecs-program-issue-brief-2018-final-pdf?inline=1
  8. Maa AY, Wojciechowski B, Hunt KJ, et al. Early experience with technology-based eye care services (TECS): a novel ophthalmologic telemedicine initiative. Ophthalmology. 2017;124(4):539-546. doi:10.1016/j.ophtha.2016.11.037
  9. Fijalkowski N, Zheng LL, Henderson MT, et al. Stanford University Network for Diagnosis of Retinopathy of Prematurity (SUNDROP): five years of screening with telemedicine. Ophthalmic Surg Lasers Imaging Retina. 2014;45(2):106-113. doi:10.3928/23258160-20140122-01
  10. Greven MA, Moshfeghi DM. Stanford University Network for Diagnosis of Retinopathy of Prematurity (SUNDROP): telemedicine-based examination after laser photocoagulation for treatment-warranted retinopathy of prematurity. Eye (Lond). 2019;33(8):1347-1355. doi:10.1038/s41433-019-0392-4
  11. US Food and Drug Administration. FDA permits marketing of artificial intelligence-based device to detect certain diabetes-related eye problems. 2018. Accessed September 28, 2021. https://www.fda.gov/news-events/press-announcements/fda-permits-marketing-artificial-intelligence-based-device-detect-certain-diabetes-related-eye
  12. Eyenuk. Eyenuk snnounces FDA clearance for EyeArt autonomous AI system for diabetic retinopathy screening. Press release. 2020. Accessed September 28, 2021. https://www.eyenuk.com/us-en/articles/diabetic-retinopathy/eyenuk-announces-eyeart-fda-clearance/2020
  13. Abramoff MD, Lavin PT, Birch M, et al. Pivotal trial of an autonomous AI-based diagnostic system for detection of diabetic retinopathy in primary care offices. NPJ Digit Med. 2018;1:39. doi:https://doi.org/10.1038/s41746-018-0040-6
  14. Schlegl T, Waldstein SM, Bogunovic H, et al. Fully automated detection and quantification of macular fluid in OCT using deep learning. Ophthalmology. 2018;125(4):549-558. doi:10.1016/j.ophtha.2017.10.031
  15. Perepelkina T, Fulton AB. Artificial intelligence (AI) applications for age-related macular degeneration (AMD) and other retinal dystrophies. Semin Ophthalmol. 2021;36(4):304-309. doi:10.1080/08820538.2021.1896756
  16. Brown JM, Campbell JP, Beers A, et al. Automated diagnosis of plus disease in retinopathy of prematurity using deep convolutional neural networks. JAMA Ophthalmol. 2018;136(7):803-810. doi:10.1001/jamaophthalmol.2018.1934
  17. Mantel I, Mosinska A, Bergin C, et al. Automated quantification of pathological fluids in neovascular age-related macular degeneration, and its repeatability using deep learning. Transl Vis Sci Technol. 2021;10(4):17. doi:10.1167/tvst.10.4.17
  18. Prahs P, Radeck V, Mayer C, et al. OCT-based deep learning algorithm for the evaluation of treatment indication with anti-vascular endothelial growth factor medications. Graefes Arch Clin Exp Ophthalmol. 2018;256(1):91-98. doi:10.1007/s00417-017-3839-y
  19. Toslak D, Ayata A, Liu C, Erol MK, Yao X. Wide-field smartphone fundus video camera based on miniaturized indirect ophthalmoscopy. Retina. 2018;38(2):438-441. doi:10.1097/IAE.0000000000001888
  20. Pieczynski J, Kuklo P, Grzybowski A. The role of telemedicine, in-home testing and artificial intelligence to alleviate an increasingly burdened healthcare system: diabetic retinopathy. Ophthalmol Ther. 2021;10(3):445-464. doi:10.1007/s40123-021-00353-2
  21. Patel TP, Kim TN, Yu G, et al. Smartphone-based, rapid, wide-field fundus photography for diagnosis of pediatric retinal diseases. Transl Vis Sci Technol. 2019;8(3):29. Published 2019 May 30. doi:10.1167/tvst.8.3.29
  22. Rajalakshmi R, Subashini R, Anjana RM, Mohan V. Automated diabetic retinopathy detection in smartphone-based fundus photography using artificial intelligence. Eye (Lond). 2018;32(6):1138-44.
  23. Toy BC, Myung DJ, He L, et al. Smartphone-based dilated fundus photography and near visual acuity testing as inexpensive screening tools to detect referral warranted diabetic eye disease. Retina. 2016;36(5):1000-1008. doi:10.1097/IAE.0000000000000955
  24. Ludwig CA, Perera C, Myung D, et al. Automatic identification of referral-warranted diabetic retinopathy using deep learning on mobile phone images. Transl Vis Sci Technol. 2020;9(2):60. Published 2020 Dec 4. doi:10.1167/tvst.9.2.60
  25. Islam M, Sansome S, Das R, et al. Smartphone-based remote monitoring of vision in macular disease enables early detection of worsening pathology and need for intravitreal therapy. BMJ Health Care Inform. 2021;28(1):e100310. doi:10.1136/bmjhci-2020-100310
  26. AREDS2-HOME Study Research Group, Chew EY, Clemons TE, et al. Randomized trial of a home monitoring system for early detection of choroidal neovascularization home monitoring of the Eye (HOME) study. Ophthalmology. 2014;121(2):535-544. doi:10.1016/j.ophtha.2013.10.027
  27. Chakravarthy U, Goldenberg D, Young G, et al. Automated identification of lesion activity in neovascular age-related macular degeneration. Ophthalmology. 2016;123(8):1731-1736. doi:10.1016/j.ophtha.2016.04.005
  28. von der Burchard C, Moltmann M, Tode J, et al. Self-examination low-cost full-field OCT (SELFF-OCT) for patients with various macular diseases. Graefes Arch Clin Exp Ophthalmol. 2021;259(6):1503-1511. doi:10.1007/s00417-020-05035-6
  29. Visotec. Our roadmap. 2021. Accessed September 28, 2021. https://visotec.health/roadmap
  30. Song G, Chu KK, Kim S, et al. First clinical application of low-cost OCT. Transl Vis Sci Technol. 2019;8(3):61. Published 2019 Jun 28. doi:10.1167/tvst.8.3.61
  31. Patel SY, Mehrotra A, Huskamp HA, Uscher-Pines L, Ganguli I, Barnett ML. Variation In telemedicine use and outpatient care during the COVID-19 pandemic in the United States. Health Aff (Millwood). 2021;40(2):349-358. doi:10.1377/hlthaff.2020.01786
  32. Aguwa UT, Aguwa CJ, Repka M, et al. Teleophthalmology in the era of COVID-19: characteristics of early adopters at a large academic institution. Telemed J E Health. 2021;27(7):739-746. doi:10.1089/tmj.2020.0372
  33. Rathi S, Tsui E, Mehta N, Zahid S, Schuman JS. The current state of teleophthalmology in the United States. Ophthalmology. 2017;124(12):1729-1734. doi:10.1016/j.ophtha.2017.05.026

What to Read Next