AUTOMATIC DETECTION OF HIGH MYOPIA AND PATHOLOGICAL MYOPIA USING FUNDUS IMAGE

AUTOMATIC DETECTION OF HIGH MYOPIA AND PATHOLOGICAL MYOPIA USING FUNDUS IMAGE

Siddharth Patil, Prof. Sweta Kale, Samrudhi Manchekar, Sankalp Manchekar and Shubham Mahalle
Savitribai Phule Pune University, RMD Sinhgad College of Engineering Warje
E-mails: siddharthmpatil1@gmail.com (Siddharth Patil), swetakale30@sinhgad.edu (Prof. Sweta Kale),

samrudhimanchekar14@gmail.com (Samrudhi Manchekar), sankalpmanchekar@gmail.com (Sankalp Manchekar), shubhammahalle195@gmail.com (Shubham Mahalle)

Abstract

Abstract—Myopia is one of the common eye defects, however it is on the increase among all global users, especially youths. Early detection and treatment are the main requirements to avoid vision loss and other complications. Pathological and high myopia detection from fundus images with the precision of "How to Develop an Automated System with Convolutional Neural Networks" is sought through this research. The proposed system in particular contains the following steps: image preprocessing, feature extraction to identify significant myopia-related features, and classification using a CNN model to classify images as either normal, pathological myopia, or high myopia. Deep learning techniques would be leveraged by the system to reach high accuracy with sensitivity in detecting myopia so as to aid healthcare professionals in the early diagnosis and timely intervention of myopia patients. This automated system would allow for a huge reduction in the screening time of myopia, particularly in a resource-poor setup.

Keywords: Pathological Myopia, High Myopia, Convolutional Neural Network