Research on diagnostic model of diabetic retinopathy based on machine learning

Jing Hou; Nan Jia; Yaoyong Duan

doi:10.1117/12.2639489

27 June 2022 Research on diagnostic model of diabetic retinopathy based on machine learning

Jing Hou, Nan Jia, Yaoyong Duan

Proceedings Volume 12253, International Conference on Automation Control, Algorithm, and Intelligent Bionics (ACAIB 2022); 1225312 (2022) https://doi.org/10.1117/12.2639489
Event: Second International Conference on Automation Control, Algorithm, and Intelligent Bionics (ACAIB 2022), 2022, Qingdao, China

Abstract

Early identification of retinopathy is of great importance in reducing blindness in diabetic patients. With the rapid development of artificial intelligence, automatic diagnosis technology of diabetic retinopathy appears. Based on the opened image data of clinical patients with diabetic retinopathy, this article establishes feature engineering through feature selection, and derives an automatic diagnosis model of retinopathy by using three machine learning algorithms: GBDT, KNN and SVM. The model is verified by real data, while the accuracy and AUC value are used to evaluate the models established by the three algorithms. Both the accuracy from ten-fold cross-validation and AUC value of GBDT are the highest, which are 0.827 and 0.803 respectively. The results show that the automatic diagnosis model of retinopathy based on GBDT algorithm has the best performance and can be an invaluable aid to the clinical diagnosis of diabetic retinopathy.

Citation Download Citation

Jing Hou, Nan Jia, and Yaoyong Duan "Research on diagnostic model of diabetic retinopathy based on machine learning", Proc. SPIE 12253, International Conference on Automation Control, Algorithm, and Intelligent Bionics (ACAIB 2022), 1225312 (27 June 2022); https://doi.org/10.1117/12.2639489

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available