In recent years, more patients were detected with corneal abnormalities. It is critical to perform diagnosis of clinical or subclinical Keratoconus accurately. The main diagnostic tool for Keratoconus is corneal topography; leading to more Laser refractive surgeries. In this paper, we recommend a novel approach which can detect Keratoconus more efficiently. The parameters derived for feature extraction are corneal volume, corneal thickness, thinnest corneal thickness, corneal area, corneal perimeter. Hybrid feed forward network along with Naive Bayes classifier is used for classification. This proposed method detects and classifies Keratoconus and the performance metrics are also evaluated using MATLAB-based simulations and finally the performance of proposed method is compared with some existing techniques.

Detection and evaluation of Keratoconus (corneal topography) by using the image classifier techniques

Computerized systems for eye maladies distinguishing proof are significant in the ophthalmology field. Conservative systems for the detection of eye disease depend on labour-intensive awareness of the retinal segments. This work exhibits another directed technique for hemorrhages discovery in advanced retinal images. This strategy utilizes an ANFIS plot for pixel association and registers a 5-D vector made out of dim dimension and Cross Section Profie (CSP) Study-constructed highlights for pixel portrayal. Classification of diseases is a crucial aspect in eye disease categorization through image processing techniques. The categorization of diseases according to pathogen groups is a significant research domain and potentially a challenging area of work. Various classification techniques for single as well as multiple diseases is identified. Classification and detection are very similar, but in classification primary focus is on the categorization of various diseases and then the classification according to various pathogen groups.

T. Sreekesh Namboodiri

Papers

Retinal Image Classification System Using Multi Phase Level Set Formulation and ANFIS