The diagnosis of diabetic retinopathy : An evaluation of different classifiers with the inception V3 model as a feature extractor

Farhan Nabil, Mohd Noor and Wan Hasbullah, Mohd Isa and Ismail, Mohd Khairuddin and Mohd Azraai, Mohd Razman and Musa, Rabiu Muazu and Ahmad Fakhri, Ab. Nasir and P.P. Abdul Majeed, Anwar (2022) The diagnosis of diabetic retinopathy : An evaluation of different classifiers with the inception V3 model as a feature extractor. In: Lecture Notes in Networks and Systems. 9th International Conference on Robot Intelligence Technology and Applications, RiTA 2021 , 16 - 17 December 2021 , Daejeon. pp. 392-397., 429 LNNS. ISSN 2367-3370 ISBN 978-303097671-2 (Published)

[img] Pdf
The diagnosis of diabetic retinopathy.pdf
Restricted to Repository staff only

Download (1MB) | Request a copy
[img]
Preview
Pdf
The diagnosis of diabetic retinopathy_An evaluation of different classifiers with the inception V3 model as a feature extractor_ABS.pdf

Download (88kB) | Preview

Abstract

Diabetic Retinopathy (DR) is a type of eye disease that is caused by diabetes mellitus. The elevated blood glucose level causes the expansion of the blood vessels that results in the leaking of the blood and other fluids. DR is a silent disease in which those inflicted with it are unaware until irregularities in the retina have advanced to the point where treatment is difficult or impossible to administer, resulting in them losing their sight completely. However, it is worth noting that early treatment can solve this problem. Hence, the purpose of this study is to develop a transfer learning pipeline for diagnosing DR. The data in the present study was obtained from the Kaggle database, and the pre-trained InceptionV3 model was employed to extract the features from the images acquired. The features are fed into the three different classifiers, namely, Support Vector Machine (SVM), k-Nearest Neighbour (kNN) and the Random Forest (RF). It was shown from the present investigation that the InceptionV3-SVM pipeline demonstrated the best performance by achieving 100%, 98% and 96% classification accuracy for the training, testing and validation dataset. The results further suggest the possible deployment of the pipeline for the diagnosis of DR.

Item Type: Conference or Workshop Item (Lecture)
Additional Information: Indexed by Scopus
Uncontrolled Keywords: Diabetic retinopathy; InceptionV3; kNN; RF; SVM; Transfer learning
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
T Technology > T Technology (General)
T Technology > TJ Mechanical engineering and machinery
T Technology > TK Electrical engineering. Electronics Nuclear engineering
T Technology > TS Manufactures
Faculty/Division: Faculty of Computing
Faculty of Manufacturing and Mechatronic Engineering Technology
Depositing User: Mr Muhamad Firdaus Janih@Jaini
Date Deposited: 30 Oct 2024 04:29
Last Modified: 30 Oct 2024 04:29
URI: http://umpir.ump.edu.my/id/eprint/42308
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item