학술논문
FundusQ-Net: A regression quality assessment deep learning algorithm for fundus images quality grading.
Document Type
Academic Journal
Author
Abramovich O; The Faculty of Biomedical Engineering, Technion-IIT, Haifa, Israel.; Pizem H; Rambam Medical Center: Rambam Health Care Campus, Israel.; Van Eijgen J; Research Group of Ophthalmology, Department of Neurosciences, KU Leuven, Oude Markt 13, 3000 Leuven; Department of Ophthalmology, University Hospitals UZ Leuven, Herestraat 49, 3000 Leuven, Belgium.; Oren I; The Faculty of Biomedical Engineering, Technion-IIT, Haifa, Israel.; Melamed J; The Faculty of Biomedical Engineering, Technion-IIT, Haifa, Israel.; Stalmans I; Research Group of Ophthalmology, Department of Neurosciences, KU Leuven, Oude Markt 13, 3000 Leuven; Department of Ophthalmology, University Hospitals UZ Leuven, Herestraat 49, 3000 Leuven, Belgium.; Blumenthal EZ; Rambam Medical Center: Rambam Health Care Campus, Israel.; Behar JA; The Faculty of Biomedical Engineering, Technion-IIT, Haifa, Israel. Electronic address: jbehar@technion.ac.il.
Source
Publisher: Elsevier Scientific Publishers Country of Publication: Ireland NLM ID: 8506513 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1872-7565 (Electronic) Linking ISSN: 01692607 NLM ISO Abbreviation: Comput Methods Programs Biomed Subsets: MEDLINE
Subject
Language
English
Abstract
Objective: Ophthalmological pathologies such as glaucoma, diabetic retinopathy and age-related macular degeneration are major causes of blindness and vision impairment. There is a need for novel decision support tools that can simplify and speed up the diagnosis of these pathologies. A key step in this process is to automatically estimate the quality of the fundus images to make sure these are interpretable by a human operator or a machine learning model. We present a novel fundus image quality scale and deep learning (DL) model that can estimate fundus image quality relative to this new scale.
Methods: A total of 1245 images were graded for quality by two ophthalmologists within the range 1-10, with a resolution of 0.5. A DL regression model was trained for fundus image quality assessment. The architecture used was Inception-V3. The model was developed using a total of 89,947 images from 6 databases, of which 1245 were labeled by the specialists and the remaining 88,702 images were used for pre-training and semi-supervised learning. The final DL model was evaluated on an internal test set (n=209) as well as an external test set (n=194).
Results: The final DL model, denoted FundusQ-Net, achieved a mean absolute error of 0.61 (0.54-0.68) on the internal test set. When evaluated as a binary classification model on the public DRIMDB database as an external test set the model obtained an accuracy of 99%.
Significance: the proposed algorithm provides a new robust tool for automated quality grading of fundus images.
Competing Interests: Declaration of Competing Interest Prof. Ingeborg Stalmans holds equities in MONA a spin off from the Catholic University of Leuven and the Belgian research institute VITO. MONA develops a solution to diagnose eye diseases from retinal pictures with artificial intelligence. The other authors have no conflicts of interests to declare.
(Copyright © 2023. Published by Elsevier B.V.)
Methods: A total of 1245 images were graded for quality by two ophthalmologists within the range 1-10, with a resolution of 0.5. A DL regression model was trained for fundus image quality assessment. The architecture used was Inception-V3. The model was developed using a total of 89,947 images from 6 databases, of which 1245 were labeled by the specialists and the remaining 88,702 images were used for pre-training and semi-supervised learning. The final DL model was evaluated on an internal test set (n=209) as well as an external test set (n=194).
Results: The final DL model, denoted FundusQ-Net, achieved a mean absolute error of 0.61 (0.54-0.68) on the internal test set. When evaluated as a binary classification model on the public DRIMDB database as an external test set the model obtained an accuracy of 99%.
Significance: the proposed algorithm provides a new robust tool for automated quality grading of fundus images.
Competing Interests: Declaration of Competing Interest Prof. Ingeborg Stalmans holds equities in MONA a spin off from the Catholic University of Leuven and the Belgian research institute VITO. MONA develops a solution to diagnose eye diseases from retinal pictures with artificial intelligence. The other authors have no conflicts of interests to declare.
(Copyright © 2023. Published by Elsevier B.V.)