Ms Rehana Khan TMA, BS Opt, FAAO

Doctoral Research Student, School of Optometry and Vision Science,
University of New South Wales, Sydney, Australia


Keywords: artificial intelligence; retina; systemic diseases 

As technology intertwines with healthcare, the eye is not only a window to the soul, but a doorway to a world of boundless possibilities. Recent years have witnessed a revolutionary transformation in ophthalmology, driven by the emergence of non-invasive, rapid, and highly precise imaging methods like retinal photography and optical coherence tomography (OCT). These techniques not only transform our assessment of ocular diseases but also unveil the eye’s macroscopic, microscopic, and molecular intricacies, illuminating aspects of both health and disease. (1)

Systemic insights through the retina

Leading this imaging revolution is the retina, a light-sensitive layer of the eye, is the only human tissue that allows a direct non-invasive visualisation of the microvascular circulation and neural tissue. It serves as a unique gateway, identifying associations between the eye and other body systems, characterising systemic diseases beyond ophthalmology.(2)

Recent studies highlight the retina’s potential in identifying systemic disease biomarkers, leading to the development of a screening tool using ocular images. The integration of artificial intelligence (AI) in this context has exhibited promising outcomes, encompassing the identification of age,(3) gender,(4) and body composition factors(5) to assessing hypertension,(6) cardiovascular disorders,(7) renal disease,(8) hepatobiliary disease,(9) and neurodegenerative diseases(10) from fundus and OCT images.

Figure 1: Distribution of articles for identifying systemic diseases from retinal Images (2)

Transparency in AI for disease detection and breaking the black box

While AI applications exhibit commendable performance in detecting systemic risk factors and diseases, their adoption faces a significant hurdle—the “black-box” nature of these algorithms. Unlike conventional statistical models, AI algorithms lack transparency, hindering the understanding of how a diagnosis is determined. This challenge has prompted the development of explanation techniques like gradient-based attribution methods and Class Activation Mapping to demystify neural network decision-making.

While these techniques offer valuable insights, they are often limited to specific classes of convolutional neural networks and restricted to image data. Essential next steps involve external validation across diverse populations to ensure the reliability and generalisability of these algorithms in detecting risk factors and diseases.

Exploring longitudinal insights beyond cross-sectional data

The extensive potential of longitudinal data remains underexplored in the development of predictive algorithms for systemic diseases. Ophthalmic imaging devices, encompassing OCT, OCT-angiography, multicolour imaging, autofluorescence and ultrawide field imaging, may present exciting opportunities to transition from cross-sectional to longitudinal perspective, providing valuable insights into disease progression for researchers and clinicians.

Real-time applications of AI in healthcare – From lab to clinical settings

As we celebrate the strides made in ophthalmic imaging, there is a pressing need for operational research to evaluate the real-time utility of AI algorithms in clinical practice. Bridging the gap between innovative technologies and everyday patient care requires rigorous assessment and validation, ensuring the seamless integration of these advancements into the ever-evolving landscape of healthcare.

In conclusion, the fusion of advanced imaging technologies and artificial intelligence holds immense promise for the future of ophthalmology and beyond. As we navigate this exciting frontier, a commitment to transparency, validation, and real-world applicability will be key to unlocking the full potential of these groundbreaking innovations.

Declaration of interest: The blog is written solely for education purpose, and it does not have any financial support and conflict of interest.



  1. Wagner SK, Fu DJ, Faes L, Liu X, Huemer J, Khalid H, Ferraz D, Korot E, Kelly C, Balaskas K, Denniston AK. Insights into systemic disease through retinal imaging-based oculomics. Translational vision science & technology. 2020 Jan 28;9(2):6-.
  2. Khan R, Surya J, Roy M, Swathi Priya MN, Mohan S, Raman S, Raman A, Vyas A, Raman R. Use of artificial intelligence algorithms to predict systemic diseases from retinal images. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery. 2023 Sep;13(5):e1506.
  3. Korot E, Pontikos N, Liu X, Wagner SK, Faes L, Huemer J, Balaskas K, Denniston AK, Khawaja A, Keane PA. Predicting sex from retinal fundus photographs using automated deep learning. Scientific reports. 2021 May 13;11(1):10286.
  4. Chueh KM, Hsieh YT, Chen HH, Ma IH, Huang SL. Identification of sex and age from macular optical coherence tomography and feature analysis using deep learning. American Journal of Ophthalmology. 2022 Mar 1;235:221-8.
  5. Gao X, Xie W, Wang Z, Zhang T, Chen B, Wang P. Predicting human body composition using a modified adaptive genetic algorithm with a novel selection operator. Plos one. 2020 Jul 16;15(7):e0235735.
  6. Zhang L, Yuan M, An Z, Zhao X, Wu H, Li H, Wang Y, Sun B, Li H, Ding S, Zeng X. Prediction of hypertension, hyperglycemia and dyslipidemia from retinal fundus photographs via deep learning: A cross-sectional study of chronic diseases in central China. PloS one. 2020 May 14;15(5):e0233166.
  7. Poplin R, Varadarajan AV, Blumer K, Liu Y, McConnell MV, Corrado GS, Peng L, Webster DR. Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nature biomedical engineering. 2018 Mar;2(3):158-64.
  8. Kang EY, Hsieh YT, Li CH, Huang YJ, Kuo CF, Kang JH, Chen KJ, Lai CC, Wu WC, Hwang YS. Deep learning–based detection of early renal function impairment using retinal fundus images: model development and validation. JMIR medical informatics. 2020 Nov 26;8(11):e23472.
  9. Xiao W, Huang X, Wang JH, Lin DR, Zhu Y, Chen C, Yang YH, Xiao J, Zhao LQ, Li JP, Cheung CY. Screening and identifying hepatobiliary diseases through deep learning using ocular images: a prospective, multicentre study. The Lancet Digital Health. 2021 Feb 1;3(2):e88-97.
  10. Satue M, Obis J, Rodrigo MJ, Otin S, Fuertes MI, Vilades E, Gracia H, Ara JR, Alarcia R, Polo V, Larrosa JM. Optical coherence tomography as a biomarker for diagnosis, progression, and prognosis of neurodegenerative diseases. Journal of Ophthalmology. 2016 Oct 20;2016.






Ms. Rehana Khan TMA is a doctoral student at the School of Optometry and Vision Science, University of New South Wales (UNSW), Sydney, Australia. Her research focuses on artificial intelligence and retinal imaging to revolutionize healthcare. Alongside her PhD, she serves as an associate lecturer and research co-supervisor at UNSW, guiding undergraduate optometry students. Recently, during her PhD internship, she took on the role of Deputy Director of Learning at the Institute of Statistical and Data Science in Australia. Rehana has completed her undergraduate (B.S. optometry) from Elite School of Optometry, in affiliation with BITS Pilani, Chennai, India and has worked as both a clinical and research optometrist at Sankara Nethralaya, Chennai. Her research interests include posterior eye segment, artificial Intelligence, deep Learning, and medical image processing. Rehana has contributed significantly to the scientific community, with numerous articles, book chapters, journal article peer reviews and has presented her work at various national and international scientific platforms. She holds certifications in programming and coding and has received grants and awards, acknowledging the significance of her research contributions.