Cover Image

EfficientNet in diabetic retinopathy detection: impact of pre-training scale

Introduction: Diabetic retinopathy (DR) is a leading cause of vision loss in diabetic patients worldwide. Timely and accurate detection of referable diabetic retinopathy (RDR) is essential for preventing blindness. In this context, deep learning technologies offer promising advancements in the autom...

Full description

Saved in:
Bibliographic Details
Main Authors: Che Azemin, Mohd Zulfaezal, Mohd Tamrin, Mohd Izzuddin, Salam, Adzura, Yusof @ Alias, Firdaus
Format: Article
Language:English
English
Published: Universiti Putra Malaysia Press 2024
Subjects:
R Medicine (General)
Online Access:http://irep.iium.edu.my/117717/7/117717_EfficientNet%20in%20diabetic%20retinopathy%20detection.pdf
http://irep.iium.edu.my/117717/13/117717_EfficientNet%20in%20diabetic%20retinopathy%20detection_Scopus.pdf
http://irep.iium.edu.my/117717/
https://medic.upm.edu.my/upload/dokumen/2024123014280818_MJMHS_0148.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
id my.iium.irep.117717
record_format dspace
spelling my.iium.irep.1177172025-01-10T02:45:44Z http://irep.iium.edu.my/117717/ EfficientNet in diabetic retinopathy detection: impact of pre-training scale Che Azemin, Mohd Zulfaezal Mohd Tamrin, Mohd Izzuddin Salam, Adzura Yusof @ Alias, Firdaus R Medicine (General) Introduction: Diabetic retinopathy (DR) is a leading cause of vision loss in diabetic patients worldwide. Timely and accurate detection of referable diabetic retinopathy (RDR) is essential for preventing blindness. In this context, deep learning technologies offer promising advancements in the automated detection of DR. This study explores the effect of pre-training dataset sizes on the diagnostic accuracy of the EfficientNet architecture for RDR identification. Methods: We utilised the EyePACS dataset containing 23,252 retinal images, including 4,343 RDR instances, for training our models. These models were then tested on the APTOS dataset, comprising 3,662 images with 1,487 cases of RDR. Two variants of EfficientNet, one pre-trained on an ImageNet-1K dataset and the other on an ImageNet-21K dataset, were compared based on their sensitivity, specificity, and AUC. Results: The EfficientNet variant pretrained on the ImageNet-1K dataset achieved a sensitivity of 97.71%, specificity of 83.13%, and an AUC of 0.901. In comparison, the variant pre-trained on the ImageNet-21K dataset demonstrated a slightly improved sensitivity of 98.79%, but a reduced specificity of 80.83%, with a comparable AUC of 0.898. Conclusion: Our study demonstrates that deep learning is a valuable tool for RDR detection, with pre-training on larger datasets resulting in modest improvements in sensitivity. However, the difference in pre-training dataset size did not substantially alter the AUC, indicating that additional factors may contribute to the overall effectiveness of the models. These results emphasize the potential of deep learning in enhancing DR screening and diagnosis, with implications for reducing the burden of diabetes-related vision loss. Malaysian Journal of Medicine and Health Sciences (2024) 20(SUPP10): 133-138. doi:10.47836/mjmhs.20. s10.18 Universiti Putra Malaysia Press 2024-12-31 Article PeerReviewed application/pdf en http://irep.iium.edu.my/117717/7/117717_EfficientNet%20in%20diabetic%20retinopathy%20detection.pdf application/pdf en http://irep.iium.edu.my/117717/13/117717_EfficientNet%20in%20diabetic%20retinopathy%20detection_Scopus.pdf Che Azemin, Mohd Zulfaezal and Mohd Tamrin, Mohd Izzuddin and Salam, Adzura and Yusof @ Alias, Firdaus (2024) EfficientNet in diabetic retinopathy detection: impact of pre-training scale. Malaysian Journal of Medicine and Health Sciences, 20 (Suppl. 10). pp. 133-138. ISSN 1675-8544 E-ISSN 2636-9346 https://medic.upm.edu.my/upload/dokumen/2024123014280818_MJMHS_0148.pdf 10.47836/mjmhs.20.s10.18
institution Universiti Islam Antarabangsa Malaysia
building IIUM Library
collection Institutional Repository
continent Asia
country Malaysia
content_provider International Islamic University Malaysia
content_source IIUM Repository (IREP)
url_provider http://irep.iium.edu.my/
language English
English
topic R Medicine (General)
spellingShingle R Medicine (General)
Che Azemin, Mohd Zulfaezal
Mohd Tamrin, Mohd Izzuddin
Salam, Adzura
Yusof @ Alias, Firdaus
EfficientNet in diabetic retinopathy detection: impact of pre-training scale
description Introduction: Diabetic retinopathy (DR) is a leading cause of vision loss in diabetic patients worldwide. Timely and accurate detection of referable diabetic retinopathy (RDR) is essential for preventing blindness. In this context, deep learning technologies offer promising advancements in the automated detection of DR. This study explores the effect of pre-training dataset sizes on the diagnostic accuracy of the EfficientNet architecture for RDR identification. Methods: We utilised the EyePACS dataset containing 23,252 retinal images, including 4,343 RDR instances, for training our models. These models were then tested on the APTOS dataset, comprising 3,662 images with 1,487 cases of RDR. Two variants of EfficientNet, one pre-trained on an ImageNet-1K dataset and the other on an ImageNet-21K dataset, were compared based on their sensitivity, specificity, and AUC. Results: The EfficientNet variant pretrained on the ImageNet-1K dataset achieved a sensitivity of 97.71%, specificity of 83.13%, and an AUC of 0.901. In comparison, the variant pre-trained on the ImageNet-21K dataset demonstrated a slightly improved sensitivity of 98.79%, but a reduced specificity of 80.83%, with a comparable AUC of 0.898. Conclusion: Our study demonstrates that deep learning is a valuable tool for RDR detection, with pre-training on larger datasets resulting in modest improvements in sensitivity. However, the difference in pre-training dataset size did not substantially alter the AUC, indicating that additional factors may contribute to the overall effectiveness of the models. These results emphasize the potential of deep learning in enhancing DR screening and diagnosis, with implications for reducing the burden of diabetes-related vision loss. Malaysian Journal of Medicine and Health Sciences (2024) 20(SUPP10): 133-138. doi:10.47836/mjmhs.20. s10.18
format Article
author Che Azemin, Mohd Zulfaezal
Mohd Tamrin, Mohd Izzuddin
Salam, Adzura
Yusof @ Alias, Firdaus
author_facet Che Azemin, Mohd Zulfaezal
Mohd Tamrin, Mohd Izzuddin
Salam, Adzura
Yusof @ Alias, Firdaus
author_sort Che Azemin, Mohd Zulfaezal
title EfficientNet in diabetic retinopathy detection: impact of pre-training scale
title_short EfficientNet in diabetic retinopathy detection: impact of pre-training scale
title_full EfficientNet in diabetic retinopathy detection: impact of pre-training scale
title_fullStr EfficientNet in diabetic retinopathy detection: impact of pre-training scale
title_full_unstemmed EfficientNet in diabetic retinopathy detection: impact of pre-training scale
title_sort efficientnet in diabetic retinopathy detection: impact of pre-training scale
publisher Universiti Putra Malaysia Press
publishDate 2024
url http://irep.iium.edu.my/117717/7/117717_EfficientNet%20in%20diabetic%20retinopathy%20detection.pdf
http://irep.iium.edu.my/117717/13/117717_EfficientNet%20in%20diabetic%20retinopathy%20detection_Scopus.pdf
http://irep.iium.edu.my/117717/
https://medic.upm.edu.my/upload/dokumen/2024123014280818_MJMHS_0148.pdf
_version_ 1821105164210143232
score 13.226497

Similar Items