Discriminative, generative artificial intelligence, and foundation models in retina imaging

Recent advances of artificial intelligence (AI) in retinal imaging found its application in two major categories: discriminative and generative AI. For discriminative tasks, conventional convolutional neural networks (CNNs) are still major AI techniques. Vision transformers (ViT), inspired by the tr...

Full description

Saved in:
Bibliographic Details
Main Authors: Paisan Ruamviboonsuk, Niracha Arjkongharn, Nattaporn Vongsa, Pawin Pakaymaskul, Natsuda Kaothanthong
Format: Article
Language:English
Published: Wolters Kluwer Medknow Publications 2024-12-01
Series:Taiwan Journal of Ophthalmology
Subjects:
Online Access:https://journals.lww.com/10.4103/tjo.TJO-D-24-00064
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850095095843389440
author Paisan Ruamviboonsuk
Niracha Arjkongharn
Nattaporn Vongsa
Pawin Pakaymaskul
Natsuda Kaothanthong
author_facet Paisan Ruamviboonsuk
Niracha Arjkongharn
Nattaporn Vongsa
Pawin Pakaymaskul
Natsuda Kaothanthong
author_sort Paisan Ruamviboonsuk
collection DOAJ
description Recent advances of artificial intelligence (AI) in retinal imaging found its application in two major categories: discriminative and generative AI. For discriminative tasks, conventional convolutional neural networks (CNNs) are still major AI techniques. Vision transformers (ViT), inspired by the transformer architecture in natural language processing, has emerged as useful techniques for discriminating retinal images. ViT can attain excellent results when pretrained at sufficient scale and transferred to specific tasks with fewer images, compared to conventional CNN. Many studies found better performance of ViT, compared to CNN, for common tasks such as diabetic retinopathy screening on color fundus photographs (CFP) and segmentation of retinal fluid on optical coherence tomography (OCT) images. Generative Adversarial Network (GAN) is the main AI technique in generative AI in retinal imaging. Novel images generated by GAN can be applied for training AI models in imbalanced or inadequate datasets. Foundation models are also recent advances in retinal imaging. They are pretrained with huge datasets, such as millions of CFP and OCT images and fine-tuned for downstream tasks with much smaller datasets. A foundation model, RETFound, which was self-supervised and found to discriminate many eye and systemic diseases better than supervised models. Large language models are foundation models that may be applied for text-related tasks, like reports of retinal angiography. Whereas AI technology moves forward fast, real-world use of AI models moves slowly, making the gap between development and deployment even wider. Strong evidence showing AI models can prevent visual loss may be required to close this gap.
format Article
id doaj-art-76cee076f2234724944eecb8914e32a4
institution DOAJ
issn 2211-5056
2211-5072
language English
publishDate 2024-12-01
publisher Wolters Kluwer Medknow Publications
record_format Article
series Taiwan Journal of Ophthalmology
spelling doaj-art-76cee076f2234724944eecb8914e32a42025-08-20T02:41:31ZengWolters Kluwer Medknow PublicationsTaiwan Journal of Ophthalmology2211-50562211-50722024-12-0114447348510.4103/tjo.TJO-D-24-00064Discriminative, generative artificial intelligence, and foundation models in retina imagingPaisan RuamviboonsukNiracha ArjkongharnNattaporn VongsaPawin PakaymaskulNatsuda KaothanthongRecent advances of artificial intelligence (AI) in retinal imaging found its application in two major categories: discriminative and generative AI. For discriminative tasks, conventional convolutional neural networks (CNNs) are still major AI techniques. Vision transformers (ViT), inspired by the transformer architecture in natural language processing, has emerged as useful techniques for discriminating retinal images. ViT can attain excellent results when pretrained at sufficient scale and transferred to specific tasks with fewer images, compared to conventional CNN. Many studies found better performance of ViT, compared to CNN, for common tasks such as diabetic retinopathy screening on color fundus photographs (CFP) and segmentation of retinal fluid on optical coherence tomography (OCT) images. Generative Adversarial Network (GAN) is the main AI technique in generative AI in retinal imaging. Novel images generated by GAN can be applied for training AI models in imbalanced or inadequate datasets. Foundation models are also recent advances in retinal imaging. They are pretrained with huge datasets, such as millions of CFP and OCT images and fine-tuned for downstream tasks with much smaller datasets. A foundation model, RETFound, which was self-supervised and found to discriminate many eye and systemic diseases better than supervised models. Large language models are foundation models that may be applied for text-related tasks, like reports of retinal angiography. Whereas AI technology moves forward fast, real-world use of AI models moves slowly, making the gap between development and deployment even wider. Strong evidence showing AI models can prevent visual loss may be required to close this gap.https://journals.lww.com/10.4103/tjo.TJO-D-24-00064discriminative artificial intelligencefoundation modelsgenerative artificial intelligenceretinal imagingvision transformer
spellingShingle Paisan Ruamviboonsuk
Niracha Arjkongharn
Nattaporn Vongsa
Pawin Pakaymaskul
Natsuda Kaothanthong
Discriminative, generative artificial intelligence, and foundation models in retina imaging
Taiwan Journal of Ophthalmology
discriminative artificial intelligence
foundation models
generative artificial intelligence
retinal imaging
vision transformer
title Discriminative, generative artificial intelligence, and foundation models in retina imaging
title_full Discriminative, generative artificial intelligence, and foundation models in retina imaging
title_fullStr Discriminative, generative artificial intelligence, and foundation models in retina imaging
title_full_unstemmed Discriminative, generative artificial intelligence, and foundation models in retina imaging
title_short Discriminative, generative artificial intelligence, and foundation models in retina imaging
title_sort discriminative generative artificial intelligence and foundation models in retina imaging
topic discriminative artificial intelligence
foundation models
generative artificial intelligence
retinal imaging
vision transformer
url https://journals.lww.com/10.4103/tjo.TJO-D-24-00064
work_keys_str_mv AT paisanruamviboonsuk discriminativegenerativeartificialintelligenceandfoundationmodelsinretinaimaging
AT nirachaarjkongharn discriminativegenerativeartificialintelligenceandfoundationmodelsinretinaimaging
AT nattapornvongsa discriminativegenerativeartificialintelligenceandfoundationmodelsinretinaimaging
AT pawinpakaymaskul discriminativegenerativeartificialintelligenceandfoundationmodelsinretinaimaging
AT natsudakaothanthong discriminativegenerativeartificialintelligenceandfoundationmodelsinretinaimaging