Smartphone-Based Fundus Photography for Diabetic Retinopathy Screening: A Scoping Review

Owais Tahhan

Introduction

Diabetes mellitus is a global epidemic, with approximately 537 million adults living with the condition, a figure projected to rise to 784 million by 2045 (1). One of its most debilitating complications is diabetic retinopathy (DR), which affects roughly one-third of the diabetic population and is a leading cause of preventable blindness (1, 2). Individuals with diabetes are 25 times more likely to become blind than the general population, making early identification of retinal microvascular damage critical for effective intervention (3, 4).

Global Burden and Early Screening

Early screening and timely treatment can prevent up to 95% of DR-related blindness cases (1, 2). However, adherence to annual screening remains low globally, often falling below 50% even in high-income countries like the United States (5, 6). In low- and middle-income countries (LMICs), access to eye care is restricted by a severe shortage of ophthalmologists and infrastructure (7, 4).

Barriers to Conventional Fundus Photography

Standard screening relies on tabletop fundus cameras, which are hindered by high costs (often exceeding $20,000), bulky design, and the requirement for highly trained personnel (5, 7). Furthermore, these devices often produce lash artifacts and peripheral distortion, and their lack of mobility makes them inaccessible to rural or bed-bound populations (8, 9).

Emergence of Smartphone-Based Fundus Photography

Smartphone-based fundus imaging (SBFI) has emerged as a portable, user-friendly, and low-cost alternative (10, 7). These devices leverage the high-resolution sensors and processing power of modern smartphones, often costing a fraction of traditional equipment—sometimes under $700 for basic adapters (2, 8).

Rationale for a Scoping Review

While SBFI is a promising technology, current evidence is heterogeneous, with studies utilizing a wide array of adapters, software, and validation protocols (11). A scoping review is necessary to map this rapidly evolving landscape and identify gaps regarding its clinical utility (2).

This scoping review aims to map the existing evidence on smartphone-based fundus photography for diabetic retinopathy screening, focusing on feasibility, diagnostic accuracy, and clinical applicability.

Methods

This scoping review was conducted to map the existing literature on the use of smartphone-based fundus photography for diabetic retinopathy screening. A systematic search was performed across PubMed/MEDLINE, Scopus, Google Scholar, and the Cochrane Library to identify relevant English-language studies.

The search strategy used combinations of keywords and Boolean operators, including “diabetic retinopathy,” “smartphone-based fundus photography,” “smartphone camera,” “handheld fundus camera,” “mobile phone,” “telemedicine,” and “screening,” with database-specific adaptations applied where necessary. Reference lists of included studies were also screened to identify additional relevant publications.

Studies were eligible if they involved adult participants (≥18 years) and evaluated smartphone-based fundus imaging devices for diabetic retinopathy screening, with comparison to established reference standards such as indirect ophthalmoscopy or conventional tabletop fundus photography.

Observational, pilot, and validation studies were included. Review articles, case reports, editorials, non-human studies, and studies focusing solely on artificial intelligence without assessment of smartphone image acquisition were excluded.

Titles and abstracts were screened for relevance, followed by full-text review of eligible studies. Data were extracted narratively, focusing on study design, imaging devices, reference standards, and screening performance. Included studies originated from diverse geographic settings, including India, Brazil, Kenya, Pakistan, Italy, and the United States, allowing assessment of the global applicability of smartphone-based fundus photography.

Discussion

Study Characteristics and Wide-field Imaging Literature identifies diverse systems, from DIY adapters using 20D lenses to commercial systems like Remidio Vistaro, D-EYE, and Peek Retina (12, 10). A significant advancement is the emergence of Wide-field Imaging (WFI); while older adapters had narrow views, devices like the Remidio Vistaro provide a 65° field of view (FOV) in a single shot (8). Through autocapture algorithms and montaging, smartphones can now achieve a 90° to 120° FOV, which exceeds the standard ETDRS seven-field cumulative FOV (8, 13).

Diagnostic Performance and AI Sophistication

SBFI shows reasonable accuracy, with AI integration significantly enhancing performance. It is important to distinguish between online (cloud-based) and offline (on-device) AI (3). Offline AI systems like Medios achieved 94% sensitivity for detecting any DR in remote field studies without internet access (3, 7). Furthermore, optimized hybrid machine learning models have achieved up to 99% accuracy on standardized datasets (12). Detection of predominantly peripheral lesions (PPL) via WFI is critical, as these lesions are associated with a 4.7-fold increased risk of DR progression (13).

Feasibility, Usability, and the Learning Curve

SBFI is highly feasible for non-ophthalmic personnel (14). Specific data on the learning curve shows that nurses with no prior experience achieved a clinical decision rate of over 80% after just 7 days of practice (14). Medical interns can acquire gradable images in 89% of eyes after only two weeks of training (5). Autocapture algorithms further simplify this, triggering automatically in 80% of examinations within 10–15 seconds (8).

Potential to Expand Access and Patient Comfort

SBFI can expand access through teleophthalmology and “mobile units,” which reduce patient travel (15, 14). In New York, a mobile unit identified new conditions in 40% of participants, many of whom had never seen an eye doctor (16). Additionally, patients often prefer SBFI due to the lower light intensity of the LED source compared to the high-intensity flash of traditional cameras (17).

Comparison and Limitations

While SBFI offers an economical workflow, limitations persist. Many devices still have a narrow field of view (20°–45°) compared to ultra-widefield tabletop systems, potentially missing lesions (13, 2). Image artifacts like lens glare remain common (18, 7). Most successful protocols still require pharmacological mydriasis, which adds time and patient discomfort (5, 13).

Future Directions

Future advancements should focus on integrating AI directly onto mobile platforms for real-time diagnosis (18). Development of “selfie fundus imaging” may eventually eliminate the need for an assistant entirely (19, 11). Additionally, mapping retinal vessel density may provide parameters similar to Optical Coherence Tomography Angiography (OCTA) using only smartphone photos (19).

Conclusion

Smartphone-based fundus photography is a reliable, portable solution for DR screening (1). Its high sensitivity and feasibility for non-specialists make it invaluable for expanding access in LMICs and underserved urban populations (14, 9). Continued innovation in AI and wide-field optics will cement its role in global blindness prevention (11, 3).

References

  1. Oliveira, J.A.E., Nakayama, L.F., Ribeiro, L.Z., Oliveira, T.V.F., Choi, S.N.J.H., Neto, E.M., Cardoso, V.S., Dib, S.A., Melo, G.B., Regatieri, C.V.S. and Malerbi, F.K. (2023) ‘Clinical validation of a smartphone-based retinal camera for diabetic retinopathy screening’, Acta Diabetologica, 60, pp. 1075–1081.
  2. Prayogo, M.E., Zaharo, A.F., Damayanti, N.N.R.R., Widyaputri, F., Thobari, J.A., Susanti, V.Y. and Sasongko, M.B. (2023) ‘Accuracy of Low-Cost, Smartphone-Based Retinal Photography for Diabetic Retinopathy Screening: A Systematic Review’, Clinical Ophthalmology, 17, pp. 2459–2470.
  3. Wroblewski, J.J., Sanchez-Buenfil, E., Inciarte, M., Berdia, J., Blake, L., Wroblewski, S., Patti, A., Suter, G. and Sanborn, G.E. (2025) ‘Diabetic Retinopathy Screening Using Smartphone-Based Fundus Photography and Deep-Learning Artificial Intelligence in the Yucatan Peninsula: A Field Study’, Journal of Diabetes Science and Technology, 19(2), pp. 370–376.
  4. Yusuf, A.M., Lusobya, R.C., Mukisa, J., Batte, C., Nakanjako, D. and Juliet-Sengeri, O. (2022) ‘Validity of smartphone-based retinal photography (PEEK-retina) compared to the standard ophthalmic fundus camera in diagnosing diabetic retinopathy in Uganda: A cross-sectional study’, PLoS ONE, 17(9), p. e0273633.
  5. Han, Y.S., Pathipati, M., Pan, C., Leung, L.-S., Blumenkranz, M.S., Myung, D. and Toy, B.C. (2021) ‘Comparison of Telemedicine Screening of Diabetic Retinopathy by Mydriatic Smartphone-Based vs Nonmydriatic Tabletop Camera-Based Fundus Imaging’, Journal of VitreoRetinal Diseases, 5(3), pp.199–207.
  6. Kim, T.N., Aaberg, M.T., Li, P., Davila, J.R., Bhaskaranand, M., Bhat, S., Ramachandra, C., Solanki, K., Myers, F., Reber, C., Jalalizadeh, R., Margolis, T.P., Fletcher, D. and Paulus, Y.M. (2021) ‘Comparison of automated and expert human grading of diabetic retinopathy using smartphone-based retinal photography’, Eye, 35, pp. 334–342.
  7. Naz, H., Nijhawan, R. and Ahuja, N.J. (2024) ‘Clinical utility of handheld fundus and smartphone-based camera for monitoring diabetic retinal diseases: a review study’, International Ophthalmology, 44, p. 41.
  8. Sivaraman, A., Nagarajan, S., Vadivel, S., Dutt, S., Tiwari, P., Narayana, S. and Rao, D.P. (2021) ‘A Novel, Smartphone-Based, Teleophthalmology-Enabled, Widefield Fundus Imaging Device With an Autocapture Algorithm’, Translational Vision Science & Technology, 10(12), p. 21.
  9. Vishwanath, Swamy, D.T. and Gaddi, D.S. (2023) ‘Smartphone for retinal imaging – Case series in resource-limited rural settings’, Indian Journal of Ophthalmology, 71, pp. 2008–2013.
  10. Jansen, L.G., Schultz, T., Holz, F.G., Finger, R.P. and Wintergerst, M.W.M. (2022) ‘Smartphone-basierte Fundusfotografie: Anwendungen und Adapter’, Der Ophthalmologe, 119, pp. 112–126.
  11. Vaughan, N. (2024) ‘Review of smartphone funduscopy for diabetic retinopathy screening’, Survey of Ophthalmology, 69, pp. 279–286.
  12. Gupta, S., Thakur, S. and Gupta, A. (2022) ‘Optimized hybrid machine learning approach for smartphone based diabetic retinopathy detection’, Multimedia Tools and Applications, 81, pp. 14475–14501.
  13. Rajalakshmi, R., Mohammed, R., Vengatesan, K., PramodKumar, T.A., Venkatesan, U., Usha, M., Arulmalar, S., Prathiba, V. and Mohan, V. (2024) ‘Wide-field imaging with smartphone based fundus camera: grading of severity of diabetic retinopathy and locating peripheral lesions in diabetic retinopathy’, Eye, 38, pp. 1471–1476.
  14. Queiroz, M.S., de Carvalho, J.X., Bortoto, S.F., de Matos, M.R., das Graças Dias Cavalcante, C., Andrade, E.A.S., Correa-Giannella, M.L. and Malerbi, F.K. (2020) ‘Diabetic retinopathy screening in urban primary care setting with a handheld smartphone-based retinal camera’, Acta Diabetologica, 57, pp. 1493–1499.
  15. Appukumran, R., Shyamsundar, K., Agrawal, M., Khurana, R., Pannu, A. and Kumar, P. (2022) ‘Eight years’ experience in mobile teleophthalmology for diabetic retinopathy screening’, Medical Hypothesis, Discovery & Innovation in Ophthalmology, 11(4), pp. 162–170.
  16. Al-Aswad, L.A., Elgin, C.Y., Patel, V., Popplewell, D., Gopal, K., Gong, D., Thomas, Z., Joiner, D., Chu, C.-K., Walters, S., Ramachandran, M., Kapoor, R., Rodriguez, M., Alcantara-Castillo, J., Maestre, G.E., Lee, J.H. and Moazami, G. (2021) ‘Real-Time Mobile Teleophthalmology for the Detection of Eye Disease in Minorities and Low Socioeconomics At-Risk Populations’, Asia-Pacific Journal of Ophthalmology (Philadelphia), 10(5), pp. 461–472.
  17. Shahzad, R., Mehmood, A., Shabbir, D. and Siddiqui, M.A.R. (2024) ‘Diagnostic accuracy of a smartphone-based device (VistaView) for detection of diabetic retinopathy: A prospective study’, PLOS Digital Health, 3(11), p. e0000649.
  18. Malerbi, F.K., Nakayama, L.F., Melo, G.B., Stuchi, J.A., Lencione, D., Prado, P.V., Ribeiro, L.Z., Dib, S.A. and Regatieri, C.V. (2024) ‘Automated Identification of Different Severity Levels of Diabetic Retinopathy Using a Handheld Fundus camera and Single-Image Protocol’, Ophthalmology Science, 4, p. 100481.
  19. Pujari, A., Saluja, G., Agarwal, D., Sinha, A., P R, A., Kumar, A. and Sharma, N. (2021) ‘Clinical Role of Smartphone Fundus Imaging in Diabetic Retinopathy and Other Neuro-retinal Diseases’, Current Eye Research, 46(11), pp. 1605–1613

Leave a Reply