Diabetic Damage Detection in Retinal Images Via a Sparsely-Connected Neurofuzzy Network

  • Authors:

  • Leonarda Carnimeo

  • Affiliations:

  • Dipartimento di Elettrotecnica ed Elettronica, Politecnico di Bari, Bari, Italy 4-70125

  • Venue:
  • ICIC '08 Proceedings of the 4th international conference on Intelligent Computing: Advanced Intelligent Computing Theories and Applications - with Aspects of Artificial Intelligence
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper a contribution towards diabetic damage detection in retinal images is proposed by synthesizing a Sparsely Connected Neurofuzzy Network for fundus image processing in the presence of retinopathies. A Hopfield-like neurofuzzy subnetwork is firstly synthesized to obtain contrast-enhanced images. After an optimal thresholding performed by an MLP-based neural subsystem, contrast-enhanced images are then globally segmented by a further sparsely-connected neural subnet to highlight vague pale regions. In this way diabetic damaged areas reveal isolated in bipolar output images. Experimental cases are reported and discussed.