Detection of diabetic retinopathy using artificial intelligence: an exploratory systematic review DOI
Richard Injante,

Marck Julca

LatIA, Journal Year: 2024, Volume and Issue: 2, P. 112 - 112

Published: Sept. 2, 2024

Diabetic retinopathy is a disease that can lead to vision loss and blindness in people with diabetes, so its early detection important prevent ocular complications. The aim of this study was analyze the usefulness artificial intelligence diabetic retinopathy. For purpose, an exploratory systematic review performed, collecting 77 empirical articles from Scopus, IEEE, ACM, SciELO NIH databases. results indicate most commonly used factors for include changes retinal vascularization, macular edema microaneurysms. Among applied algorithms are ResNet 101, CNN IDx-DR. In addition, some models reported have accuracy ranging 90% 95%, although accuracies below 80% also been identified. It concluded intelligence, particular deep learning, has shown be effective retinopathy, facilitating timely treatment improving clinical outcomes. However, ethical legal concerns arise, such as privacy security patient data, liability case diagnostic errors, algorithmic bias, informed consent, transparency use intelligence.

Language: Английский

Understanding natural language: Potential application of large language models to ophthalmology DOI Creative Commons
Zefeng Yang, Biao Wang, Fengqi Zhou

et al.

Asia-Pacific Journal of Ophthalmology, Journal Year: 2024, Volume and Issue: 13(4), P. 100085 - 100085

Published: July 1, 2024

Large language models (LLMs), a natural processing technology based on deep learning, are currently in the spotlight. These closely mimic comprehension and generation. Their evolution has undergone several waves of innovation similar to convolutional neural networks. The transformer architecture advancement generative artificial intelligence marks monumental leap beyond early-stage pattern recognition via supervised learning. With expansion parameters training data (terabytes), LLMs unveil remarkable human interactivity, encompassing capabilities such as memory retention comprehension. advances make particularly well-suited for roles healthcare communication between medical practitioners patients. In this comprehensive review, we discuss trajectory their potential implications clinicians For clinicians, can be used automated documentation, given better inputs extensive validation, may able autonomously diagnose treat future. patient care, triage suggestions, summarization documents, explanation patient's condition, customizing education materials tailored level. limitations possible solutions real-world use also presented. Given rapid advancements area, review attempts briefly cover many that play ophthalmic space, with focus improving quality delivery.

Language: Английский

Citations

7
Artificial Intelligence Applications in Diabetic Retinopathy: What We Have Now and What to Expect in the Future DOI Creative Commons
Mingui Kong, Su Jeong Song

Endocrinology and Metabolism, Journal Year: 2024, Volume and Issue: 39(3), P. 416 - 424

Published: June 10, 2024

Diabetic retinopathy (DR) is a major complication of diabetes mellitus and leading cause vision loss globally. A prompt accurate diagnosis crucial for ensuring favorable visual outcomes, highlighting the need increased access to medical care. The recent remarkable advancements in artificial intelligence (AI) have raised high expectations its role disease prognosis prediction across various fields. In addition achieving precision comparable that ophthalmologists, AI-based DR has potential improve accessibility, especially through telemedicine. this review paper, we aim examine current AI explore future directions.

Language: Английский

Citations

5
Applications of ChatGPT in the diagnosis, management, education, and research of retinal diseases: a scoping review DOI Creative Commons
Victor C. F. Bellanda, Manuel Filipe Santos, Daniel Ferraz

et al.

International Journal of Retina and Vitreous, Journal Year: 2024, Volume and Issue: 10(1)

Published: Oct. 17, 2024

This scoping review aims to explore the current applications of ChatGPT in retina field, highlighting its potential, challenges, and limitations.

Language: Английский

Citations

4
Prospects and perils of ChatGPT in diabetes DOI
GR Sridhar,

Lakshmi Gumpeny

World Journal of Diabetes, Journal Year: 2025, Volume and Issue: 16(3)

Published: Jan. 20, 2025

ChatGPT, a popular large language model developed by OpenAI, has the potential to transform management of diabetes mellitus. It is conversational artificial intelligence trained on extensive datasets, although not specifically health-related. The development and core components ChatGPT include neural networks machine learning. Since current yet diabetes-related it limitations such as risk inaccuracies need for human supervision. Nevertheless, aid in patient engagement, medical education, clinical decision support. In management, can contribute personalized dietary guidelines, providing emotional Specifically, being tested scenarios assessment obesity, screening diabetic retinopathy, provision guidelines ketoacidosis. Ethical legal considerations are essential before be integrated into healthcare. Potential concerns relate data privacy, accuracy responses, maintenance patient-doctor relationship. Ultimately, while models hold immense revolutionize care, one needs weigh their limitations, ethical implications, integration promises future proactive, personalized, patient-centric care management.

Language: Английский

Citations

0
Accuracy of Artificial Intelligence Versus Clinicians in Real-Life Case Scenarios of Retinopathy of Prematurity DOI Open Access
Akash Belenje,

Devesh M. Pandya,

Subhadra Jalali

et al.

Cureus, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 5, 2025

Objective The aim of this study was to compare the accuracy ChatGPT artificial intelligence (AI) with clinicians in real-life case scenarios related retinopathy prematurity (ROP). Methods This a prospectively conducted scenario-based questionnaire multiple-response answers. Thirteen clinicians, including eight vitreoretinal fellowship trainees (with less than two years experience management ROP) and five ROP experts more three ROP), were given 10 ROP. majority responses from compared AI-generated responses. exercise repeated for both versions 3.5 4.0 month apart on May 29, 2024, July 18, check AI response consistency. For each scenario, clinician agreement. Results answered nine cases correctly (90%), outperforming (77.5%, i.e., 62 correct out 80). highest at 96% (i.e., 48 50). There substantial agreement between responses, Cohen's kappa 0.80. Conclusion model showed performed better trainees. presents promising new software tools that can be explored further use A accurate prompt mentioning type screening guidelines promote answers by as per requested guidelines.

Language: Английский

Citations

0
Vision of the future: large language models in ophthalmology DOI
Prashant D. Tailor, Haley S. D’Souza, Hanzhou Li

et al.

Current Opinion in Ophthalmology, Journal Year: 2024, Volume and Issue: 35(5), P. 391 - 402

Published: May 30, 2024

Large language models (LLMs) are rapidly entering the landscape of medicine in areas from patient interaction to clinical decision-making. This review discusses evolving role LLMs ophthalmology, focusing on their current applications and future potential enhancing ophthalmic care.

Language: Английский

Citations

3
Discriminative, generative artificial intelligence, and foundation models in retina imaging DOI Creative Commons
Paisan Ruamviboonsuk, Niracha Arjkongharn, Nattaporn Vongsa

et al.

Taiwan Journal of Ophthalmology, Journal Year: 2024, Volume and Issue: 14(4), P. 473 - 485

Published: Oct. 1, 2024

Abstract Recent advances of artificial intelligence (AI) in retinal imaging found its application two major categories: discriminative and generative AI. For tasks, conventional convolutional neural networks (CNNs) are still AI techniques. Vision transformers (ViT), inspired by the transformer architecture natural language processing, has emerged as useful techniques for discriminating images. ViT can attain excellent results when pretrained at sufficient scale transferred to specific tasks with fewer images, compared CNN. Many studies better performance ViT, CNN, common such diabetic retinopathy screening on color fundus photographs (CFP) segmentation fluid optical coherence tomography (OCT) Generative Adversarial Network (GAN) is main technique imaging. Novel images generated GAN be applied training models imbalanced or inadequate datasets. Foundation also recent They huge datasets, millions CFP OCT fine-tuned downstream much smaller A foundation model, RETFound, which was self-supervised discriminate many eye systemic diseases than supervised models. Large that may text-related like reports angiography. Whereas technology moves forward fast, real-world use slowly, making gap between development deployment even wider. Strong evidence showing prevent visual loss required close this gap.

Language: Английский

Citations

2
Evaluation of Systemic Risk Factors in Patients with Diabetes Mellitus for Detecting Diabetic Retinopathy with Random Forest Classification Model DOI Creative Commons
Ramesh Venkatesh, Priyanka Gandhi, Ayushi Choudhary

et al.

Diagnostics, Journal Year: 2024, Volume and Issue: 14(16), P. 1765 - 1765

Published: Aug. 13, 2024

Background: This study aims to assess systemic risk factors in diabetes mellitus (DM) patients and predict diabetic retinopathy (DR) using a Random Forest (RF) classification model. Methods: We included DM presenting the retina clinic for first-time DR screening. Data on age, gender, type, treatment history, control status, family pregnancy comorbidities were collected. sight-threatening (STDR) diagnosed via dilated fundus examination. The dataset was split 80:20 into training testing sets. RF model trained detect STDR separately, its performance evaluated misclassification rates, sensitivity, specificity. Results: from 1416 analyzed. 1132 (80%) patients. rates 0% ~20% set. External 284 (20%) showed 100% accuracy, specificity detection. For STDR, achieved 76% (95% CI-70.7%–80.7%) 53% CI-39.2%–66.6%) 80% CI-74.6%–84.7%) Conclusions: effectively predicts factors, potentially reducing unnecessary referrals However, further validation with diverse datasets is necessary establish reliability clinical use.

Language: Английский

Citations

1
Detection of diabetic retinopathy using artificial intelligence: an exploratory systematic review DOI
Richard Injante,

Marck Julca

LatIA, Journal Year: 2024, Volume and Issue: 2, P. 112 - 112

Published: Sept. 2, 2024

Diabetic retinopathy is a disease that can lead to vision loss and blindness in people with diabetes, so its early detection important prevent ocular complications. The aim of this study was analyze the usefulness artificial intelligence diabetic retinopathy. For purpose, an exploratory systematic review performed, collecting 77 empirical articles from Scopus, IEEE, ACM, SciELO NIH databases. results indicate most commonly used factors for include changes retinal vascularization, macular edema microaneurysms. Among applied algorithms are ResNet 101, CNN IDx-DR. In addition, some models reported have accuracy ranging 90% 95%, although accuracies below 80% also been identified. It concluded intelligence, particular deep learning, has shown be effective retinopathy, facilitating timely treatment improving clinical outcomes. However, ethical legal concerns arise, such as privacy security patient data, liability case diagnostic errors, algorithmic bias, informed consent, transparency use intelligence.

Language: Английский

Citations

1