Grading Severity of Pterygium using Fuzzy Reasoning

Abstract

Reliable and accurate severity pronouncements are essential for clinical and epidemiologic related maladies. This paper presents the development of automated system that would detect and assess the damage caused due to pterygium growth. The development of the system included 2 distinct stages. In first stage a basic system is developed which could measure the 3 features of any input image (containing pterygium occurring in corneal region) and in the second stage assessment of damage has been done using Fuzzy Inference System. Most of the researchers have considered the extent of pterygium in terms of linear measures. Here, in this work the Redness of pterygium is established has a novel feature that could indicate the growth tendency of pterygium. Among many soft computing techniques Fuzzy Inference System (FIS) proved to be accurate to the extent of 91.4% accuracy. The other parameters like specificity, sensitivity are also adequate for accurate assessment of damage caused by pterygium