Conference Proceeding

Image Quality Classification for DR Screening Using Convolutional Neural Networks

  • Ruwan Tennakoon (IBM Research Australia)
  • Dwarikanath Mahapatra (IBM Research Australia)
  • Pallab Roy (IBM Research Australia)
  • Suman Sedai (IBM Research Australia)
  • Rahil Garnavi (IBM Research Australia)


The quality of input images significantly affects the outcome of automated diabetic retinopathy screening systems. Current methods to identify image quality rely on hand-crafted geometric and structural features, that does not generalize well. We propose a new method for retinal image quality classification (IQC) that uses computational algorithms imitating the working of the human visual systems. The proposed method leverages on learned supervised information using convolutional neural networks (CNN), thus avoiding hand-engineered features. Our analysis shows that the learned features capture both geometric and structural information relevant for image quality classification. Experimental results conducted on a relatively large dataset demonstrates that the overall method can achieve high accuracy. We also show that effective features for IQC can be learned by full training of shallow CNN as well as by using transfer learning.

Keywords: Digital fundus images, Diabetic retinopathy, Image quality classification, CNN, retinal imaging

How to Cite:

Tennakoon, R. & Mahapatra, D. & Roy, P. & Sedai, S. & Garnavi, R., (2016) “Image Quality Classification for DR Screening Using Convolutional Neural Networks”, Proceedings of the Ophthalmic Medical Image Analysis International Workshop 3(2016), 113-120. doi:

Rights: Copyright © 2016 the authors

Download pdf
View PDF



Published on
21 Oct 2016
Peer Reviewed