FundusQ-Net: A regression quality assessment deep learning algorithm for fundus images quality grading

• Published in: Computer methods and programs in biomedicine Vol. 239; p. 107522
• Main Authors: Abramovich, Or, Pizem, Hadas, Van Eijgen, Jan, Oren, Ilan, Melamed, Joshua, Stalmans, Ingeborg, Blumenthal, Eytan Z., Behar, Joachim A.
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01.09.2023
• Subjects: Deep learning; Fundus image; Quality assessment; Semi supervised learning; Fundus image; Deep learning; Quality assessment; Semi supervised learning
• ISSN: 0169-2607; 1872-7565; 1872-7565
• DOI: 10.1016/j.cmpb.2023.107522

•There is a lack of computational tools for quality estimation of fundus images.•We introduce a novel fundus image quality scale (range 1–10).•We develop a new deep learning model, FundusQ-Net, that can estimate fundus image quality relative to this new scale.•FundusQ-Net is trained on 89,947 images and demonstrates high precision and to generalize to an external dataset. 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.