We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Image and Video Processing

Title: FundusQ-Net: a Regression Quality Assessment Deep Learning Algorithm for Fundus Images Quality Grading

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 1,245 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 1,245 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.
Comments: 7 pages, 7 figures, submitted to Computer Methods and Programs in Biomedicine
Subjects: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
ACM classes: I.2.10
Cite as: arXiv:2205.01676 [eess.IV]
  (or arXiv:2205.01676v1 [eess.IV] for this version)

Submission history

From: Or Abramovich [view email]
[v1] Mon, 2 May 2022 21:01:34 GMT (1528kb,D)

Link back to: arXiv, form interface, contact.