Evaluating deep learning models for classifying OCT images with limited data and noisy labels

Abstract The use of deep learning for OCT image classification could enhance the diagnosis and monitoring of retinal diseases. However, challenges like variability in retinal abnormalities, noise, and artifacts in OCT images limit its clinical use. Our study aimed to evaluate the performance of vari...

Full description

Saved in:
Bibliographic Details
Main Authors: Aleksandar Miladinović, Alessandro Biscontin, Miloš Ajčević, Simone Kresevic, Agostino Accardo, Dario Marangoni, Daniele Tognetto, Leandro Inferrera
Format: Article
Language:English
Published: Nature Portfolio 2024-12-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-024-81127-1
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846137084068233216
author Aleksandar Miladinović
Alessandro Biscontin
Miloš Ajčević
Simone Kresevic
Agostino Accardo
Dario Marangoni
Daniele Tognetto
Leandro Inferrera
author_facet Aleksandar Miladinović
Alessandro Biscontin
Miloš Ajčević
Simone Kresevic
Agostino Accardo
Dario Marangoni
Daniele Tognetto
Leandro Inferrera
author_sort Aleksandar Miladinović
collection DOAJ
description Abstract The use of deep learning for OCT image classification could enhance the diagnosis and monitoring of retinal diseases. However, challenges like variability in retinal abnormalities, noise, and artifacts in OCT images limit its clinical use. Our study aimed to evaluate the performance of various deep learning (DL) architectures in classifying retinal pathologies versus healthy cases based on OCT images, under data scarcity and label noise. We examined five DL architectures: ResNet18, ResNet34, ResNet50, VGG16, and InceptionV3. Fine-tuning of the pre-trained models was conducted on 5526 OCT images and reduced subsets down to 21 images to evaluate performance under data scarcity. The performance of models fine-tuned on subsets with label noise levels of 10%, 15%, and 20% was evaluated. All DL architectures achieved high classification accuracy (> 90%) with training sets of 345 or more images. InceptionV3 achieved the highest classification accuracy (99%) when trained on the entire training set. However, classification accuracy decreased and variability increased as sample size decreased. Label noise significantly affected model accuracy. Compensating for labeling errors of 10%, 15%, and 20% requires approximately 4, 9, and 14 times more images in the training set to reach the performance of 345 correctly labeled images. The results showed that DL models fine-tuned on sets of 345 or more OCT images can accurately classify retinal pathologies versus healthy controls. Our findings highlight that while mislabeling errors significantly impact classification performance in OCT analysis, this can be effectively mitigated by increasing the training sample size. By addressing data scarcity and labeling errors, our research aims to improve the real-world application and accuracy of retinal disease management.
format Article
id doaj-art-3f356192cf4c49c78321c6cc81b3adc9
institution Kabale University
issn 2045-2322
language English
publishDate 2024-12-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-3f356192cf4c49c78321c6cc81b3adc92024-12-08T12:29:44ZengNature PortfolioScientific Reports2045-23222024-12-0114111110.1038/s41598-024-81127-1Evaluating deep learning models for classifying OCT images with limited data and noisy labelsAleksandar Miladinović0Alessandro Biscontin1Miloš Ajčević2Simone Kresevic3Agostino Accardo4Dario Marangoni5Daniele Tognetto6Leandro Inferrera7Institute for Maternal and Child Health IRCCS “Burlo Garofolo”Institute for Maternal and Child Health IRCCS “Burlo Garofolo”Department of Engineering and Architecture, University of TriesteDepartment of Engineering and Architecture, University of TriesteDepartment of Engineering and Architecture, University of TriesteDepartment of Medicine, Surgery and Health Sciences, Ophthalmology Clinic, University of TriesteDepartment of Medicine, Surgery and Health Sciences, Ophthalmology Clinic, University of TriesteDepartment of Medicine, Surgery and Health Sciences, Ophthalmology Clinic, University of TriesteAbstract The use of deep learning for OCT image classification could enhance the diagnosis and monitoring of retinal diseases. However, challenges like variability in retinal abnormalities, noise, and artifacts in OCT images limit its clinical use. Our study aimed to evaluate the performance of various deep learning (DL) architectures in classifying retinal pathologies versus healthy cases based on OCT images, under data scarcity and label noise. We examined five DL architectures: ResNet18, ResNet34, ResNet50, VGG16, and InceptionV3. Fine-tuning of the pre-trained models was conducted on 5526 OCT images and reduced subsets down to 21 images to evaluate performance under data scarcity. The performance of models fine-tuned on subsets with label noise levels of 10%, 15%, and 20% was evaluated. All DL architectures achieved high classification accuracy (> 90%) with training sets of 345 or more images. InceptionV3 achieved the highest classification accuracy (99%) when trained on the entire training set. However, classification accuracy decreased and variability increased as sample size decreased. Label noise significantly affected model accuracy. Compensating for labeling errors of 10%, 15%, and 20% requires approximately 4, 9, and 14 times more images in the training set to reach the performance of 345 correctly labeled images. The results showed that DL models fine-tuned on sets of 345 or more OCT images can accurately classify retinal pathologies versus healthy controls. Our findings highlight that while mislabeling errors significantly impact classification performance in OCT analysis, this can be effectively mitigated by increasing the training sample size. By addressing data scarcity and labeling errors, our research aims to improve the real-world application and accuracy of retinal disease management.https://doi.org/10.1038/s41598-024-81127-1OCTCNNDeep learningRetinaRetinal imaging
spellingShingle Aleksandar Miladinović
Alessandro Biscontin
Miloš Ajčević
Simone Kresevic
Agostino Accardo
Dario Marangoni
Daniele Tognetto
Leandro Inferrera
Evaluating deep learning models for classifying OCT images with limited data and noisy labels
Scientific Reports
OCT
CNN
Deep learning
Retina
Retinal imaging
title Evaluating deep learning models for classifying OCT images with limited data and noisy labels
title_full Evaluating deep learning models for classifying OCT images with limited data and noisy labels
title_fullStr Evaluating deep learning models for classifying OCT images with limited data and noisy labels
title_full_unstemmed Evaluating deep learning models for classifying OCT images with limited data and noisy labels
title_short Evaluating deep learning models for classifying OCT images with limited data and noisy labels
title_sort evaluating deep learning models for classifying oct images with limited data and noisy labels
topic OCT
CNN
Deep learning
Retina
Retinal imaging
url https://doi.org/10.1038/s41598-024-81127-1
work_keys_str_mv AT aleksandarmiladinovic evaluatingdeeplearningmodelsforclassifyingoctimageswithlimiteddataandnoisylabels
AT alessandrobiscontin evaluatingdeeplearningmodelsforclassifyingoctimageswithlimiteddataandnoisylabels
AT milosajcevic evaluatingdeeplearningmodelsforclassifyingoctimageswithlimiteddataandnoisylabels
AT simonekresevic evaluatingdeeplearningmodelsforclassifyingoctimageswithlimiteddataandnoisylabels
AT agostinoaccardo evaluatingdeeplearningmodelsforclassifyingoctimageswithlimiteddataandnoisylabels
AT dariomarangoni evaluatingdeeplearningmodelsforclassifyingoctimageswithlimiteddataandnoisylabels
AT danieletognetto evaluatingdeeplearningmodelsforclassifyingoctimageswithlimiteddataandnoisylabels
AT leandroinferrera evaluatingdeeplearningmodelsforclassifyingoctimageswithlimiteddataandnoisylabels