AI-powered oculomics for early diagnosis of neurodegenerative diseases through retinal microvasculature analysis - Letter to the Editor

Sep 29, 2026

AI-powered oculomics leverages advanced machine learning and deep learning to analyze retinal microvasculature for early detection of neurodegenerative diseases. The letter highlights retinal links to Alzheimer’s disease and Parkinson’s disease, including thinning of retinal nerve fiber layers, reduced perfusion density, and vascularity changes. Automated AI segmentation replaces time-consuming manual methods, and reported models such as CNN variants achieve measurable discrimination performance. Key limitations include limited datasets and insufficient generalizability, alongside the need for interdisciplinary ophthalmology-neurology collaboration.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages2

Reading Time5 min

AccessFree to download

Overview

AI methods and diagnostic evidence

Challenges and limitations

Conclusion

Letter to the Editor AI-powered oculomics for early diagnosis of neurodegenerative diseases through retinal microvasculature analysis Fakiha Rahman, MBBSa , Amna Rahman, MBBSb , Muhammad Talha, MBBSa , Noor Un Nisa Irshad, MBBSc , Sakan Binte Imran, MBBSd ,* To the Editor, Neurodegenerative diseases are identified by progressive loss of neuronal function in specific areas of the brain. Alzheimer’s disease and Parkinson’s disease are the most prevalent neurodegenerative disorders, with AD being diagnosed by memory impairment and PD being manifested by loss of motor function[1] . These diseases are found to be associated with the pathological changes in the retina. Patients with Alzheimer’s disease clinically present with weakening of retinal vasculature, thinning in the retinal nerve fiber layer, and the ganglion cellinner plexiform layer. A decrease in retinal perfusion density, vessel density, and choroidal vascularity index has been observed in patients with Parkinson’s disease. The time-consuming manual segmentation of the vasculature is now replaced with the use of advanced artificial intelligence (AI)-based models that aid in the vasculature segmentation and particularly in the decision making of ophthalmologists for the ultimate diagnosis[2,3] . This article aligns with the TITAN Guidelines on the need for transparency in AI use in healthcare[4] . The application of AI-powered oculomics, such as deep learning and machine learning, has enabled the prediction of multiple neurodegenerative diseases. These diagnostic measures from AI allow the initiation of early preventive measures that reduce morbidity and mortality[5] . The onset of more severe symptoms can also be avoided with medications in the early stages. Retina, being a valuable window into the brain, can reflect neurodegenerative processes. AD is associated with thinning of the retinal nerve fiber layer and reduced contrast sensitivity due to loss of dopaminergic retinal cells, which has been observed in PD. Asa rapidly emerging transformative technology, AI has the aDepartment of Medicine, King Edward Medical University, Lahore, Pakistan, bDepartment of Medicine, Faisalabad Medical University, Faisalabad, Pakistan, cDepartment of Medicine, Jinnah Sindh Medical University, Karachi, Pakistan and dDepartment of Medicine, Sir Salimullah Medical College, Dhaka, Bangladesh Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. *Corresponding author. Address: Sir Salimullah Medical College, Dhaka 1100, [Bangladesh. E-mail: sakanbinteimran.ssmc@gmail.com](Bangladesh. E-mail: sakanbinteimran.ssmc@gmail.com) (S.B. Imran). Copyright © 2025 The Author(s). Published by Wolters Kluwer Health, Inc. This is an open access article distributed under the Creative Commons AttributionNoDerivatives License 4.0, which allows for redistribution, commercial and noncommercial, as long as it is passed along unchanged and in whole, with credit to the author. Annals of Medicine & Surgery (2026) 88:1036–1037 Received 7 October 2025; Accepted 11 November 2025 Published online 22 November 2025 [http://dx.doi.org/10.1097/MS9.0000000000004391](http://dx.doi.org/10.1097/MS9.0000000000004391) potential for early detection of neurodegenerative diseases with a noninvasive screening method[1] . Recent clinical studies highlighted that AI algorithms focusing on retinal microvasculature identify early, minimal changes, which are associated with the diagnosis of Parkinson’s or Alzheimer’s disease. A study conducted on 615 participants showed that Convolutional Neural Network (CNN), a type of AI, successfully distinguished groups with Parkinson’s disease, with Resnet-18 being the best, with an AUROC of 0.76 and an accuracy of 76.75%[6]. Another study revealed that an area under the curve of 0.918 was achieved with CNN[7] . When AI models are combined, their efficacy increases further, as a study demonstrated that a trilate

FAQs

How does AI-powered oculomics support early diagnosis of neurodegenerative diseases?

It uses AI models to analyze retinal microvasculature and associated retinal changes linked to diseases such as Alzheimer’s and Parkinson’s, enabling earlier identification and supporting ophthalmologists’ decision-making.

Which AI approaches are mentioned as effective for retinal-based disease discrimination?

The letter cites deep learning and machine learning, including convolutional neural networks (CNN) and specific variants such as Resnet-18, as well as combined models and explainable AI frameworks like LAVA.

What challenges limit the clinical generalizability of current AI models?

Models were trained on limited datasets with the same data used for training, validation, and testing, reducing general applicability; the letter also notes that single-parameter models may not be clinically trustworthy without upgrades and stronger validation setups.

Sep 29, 2026

Upload by CatatanPagi