Citation: Tsai, C.-Y.; Chen, C.-T.;
Chen, G.-A.; Yeh, C.-F.; Kuo, C.-T.;
Hsiao, Y.-C.; Hu, H.-Y.; Tsai, I.-L.;
Wang, C.-H.; Chen, J.-R.; et al.
Necessity of Local Modification for
Deep Learning Algorithms to Predict
Diabetic Retinopathy. Int. J. Environ.
Res. Public Health 2022, 19, 1204.
https://doi.org/10.3390/
ijerph19031204
Academic Editors: Keun Ho Ryu and
Nipon Theera-Umpon
Received: 17 December 2021
Accepted: 18 January 2022
Published: 21 January 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
International Journal of
Environmental Research
and Public Health
Article
Necessity of Local Modification for Deep Learning Algorithms
to Predict Diabetic Retinopathy
Ching-Yao Tsai
1,2,3
, Chueh-Tan Chen
1,4
, Guan-An Chen
5
, Chun-Fu Yeh
5
, Chin-Tzu Kuo
1
, Ya-Chuan Hsiao
1,6
,
Hsiao-Yun Hu
2,7,8
, I-Lun Tsai
1
, Ching-Hui Wang
1
, Jian-Ren Chen
5
, Su-Chen Huang
5
, Tzu-Chieh Lu
5
and Lin-Chung Woung
1,9,10,
*
1
Department of Ophthalmology, Taipei City Hospital, Taipei 103, Taiwan; dac58@tpech.gov.tw (C.-Y.T.);
chuehtan@hotmail.com (C.-T.C.); sbodys@yahoo.com.tw (C.-T.K.); daj20daj20@gmail.com (Y.-C.H.);
ilunt@ms49.hinet.net (I-L.T.); B0493@tpech.gov.tw (C.-H.W.)
2
Institute of Public Health, National Yang Ming Chiao Tung University, Taipei 112, Taiwan;
A3547@tpech.gov.tw
3
Department of Business Administration, Fu Jen Catholic University, New Taipei City 242, Taiwan
4
Institute of Traditional Medicine, National Yang Ming Chiao Tung University, Taipei 112, Taiwan
5
Smart Medical and Healthcare, Service Systems Technology Center, Industrial Technology Research Institute,
Hsinchu County 310, Taiwan; crackereidolon@gmail.com (G.-A.C.); r02429007@ntu.edu.tw (C.-F.Y.);
cjr@itri.org.tw (J.-R.C.); schuang@itri.org.tw (S.-C.H.); OrionLu@itri.org.tw (T.-C.L.)
6
College of Medicine, National Yang Ming Chiao Tung University, Taipei 112, Taiwan
7
Department of Education and Research, Taipei City Hospital, Taipei 106, Taiwan
8
Department of Health and Welfare, University of Taipei, Taipei 100, Taiwan
9
Department of Health Care Management, National Taipei University of Nursing and Health Sciences,
Taipei 112, Taiwan
10
Department of Medicine, National Yang Ming Chiao Tung University, Taipei 112, Taiwan
* Correspondence: ayitgroup@gmail.com; Tel.: +886-2-2552-3234
Abstract:
Deep learning (DL) algorithms are used to diagnose diabetic retinopathy (DR). However,
most of these algorithms have been trained using global data or data from patients of a single
region. Using different model architectures (e.g., Inception-v3, ResNet101, and DenseNet121), we
assessed the necessity of modifying the algorithms for universal society screening. We used the
open-source dataset from the Kaggle Diabetic Retinopathy Detection competition to develop a model
for the detection of DR severity. We used a local dataset from Taipei City Hospital to verify the
necessity of model localization and validated the three aforementioned models with local datasets.
The experimental results revealed that Inception-v3 outperformed ResNet101 and DenseNet121 with
a foreign global dataset, whereas DenseNet121 outperformed Inception-v3 and ResNet101 with the
local dataset. The quadratic weighted kappa score (
κ
) was used to evaluate the model performance.
All models had 5–8% higher
κ
for the local dataset than for the foreign dataset. Confusion matrix
analysis revealed that, compared with the local ophthalmologists’ diagnoses, the severity predicted
by the three models was overestimated. Thus, DL algorithms using artificial intelligence based on
global data must be locally modified to ensure the applicability of a well-trained model to make
diagnoses in clinical environments.
Keywords: diabetic retinopathy; deep learning algorithms; model localised; Taiwan; predict
1. Introduction
Diabetic retinopathy is one of the leading causes of blindness worldwide. However,
there are no specific symptoms of early diabetic retinopathy, which results in both delayed
diagnosis and disease progression in diabetic patients. Thus, the popularity of deep
learning algorithms predicting vision-threatening diabetic retinopathy is arising. In recent
years, deep learning has achieved great success in medical image analysis. However, most
works directly employ algorithms based on convolutional neural networks (CNNs), which
Int. J. Environ. Res. Public Health 2022, 19, 1204. https://doi.org/10.3390/ijerph19031204 https://www.mdpi.com/journal/ijerph
