Detection and differentiation of retinal exudates in color retinal images

Authors

  • Germán Sánchez Torres Universidad del Magdalena
  • Jorge Rudas Castaño Universidad del Magdalena
  • Ricardo Toscano Cuello Universidad del Magdalena

Abstract

Diabetic retinopathy early is not usually perceivable in patients with diabetic. Their first signs, like micro-aneurysms, often are unnoticed in preliminary test due to its difficult detection by means specialists. In addition, its presence is necessarily not directly related to diabetic retinopathy, since there are others pathologies that may also cause those. Until the detection of the presence of exudates, a specialist can deduce the presence of diabetic retinopathy. This paper presents a method to assist the identification and differentiation of exudates on color retinal images based on a variety of k-nn filters. The proposed method proved to be a sensitive and specific detection of bright lesions, able to determine with sufficient certainty, a possible injury, with a specificity of 99%.

Author Biographies

Germán Sánchez Torres, Universidad del Magdalena

Docente Universidad del Magdalena, Facultad de Ingeniería. Magister en Ingeniería de Sistemas, Universidad Nacional de Colombia.

Jorge Rudas Castaño, Universidad del Magdalena

Ingeniero de Sistemas, Universidad del Magdalena (Colombia).

Ricardo Toscano Cuello, Universidad del Magdalena

Ingeniero de Sistemas, Universidad del Magdalena (Colombia).

Published

2011-12-11

How to Cite

[1]
G. Sánchez Torres, J. Rudas Castaño, and R. Toscano Cuello, “Detection and differentiation of retinal exudates in color retinal images”, Ing. y Des., vol. 29, no. 2, pp. 224–241, Dec. 2011.