Webová aplikace pro computer-assisted diagnostiku retinálních záznamů

Abstract

The most common cause of vision impairment and loss in premature babies is a disease known as Retinopathy of prematurity. To detect the occurrence, retinal imaging of these babies is done as a preventive measure immediately after birth, which can be very stressful. Thus, the prevailing effort is to eliminate this method. These images are then analyzed by an opthamologist who determines susceptibility to the disease based on retinal artifacts. However, it is often difficult to spot these artefacts, especially for novice ophthamologists. This bachelor thesis deals with the development of a web application for displaying, filtering and transforming these images. Its aim is to provide the wider public, and possibly doctors in the field, with the ability to manage the captured images in one place and apply image transformations to them to highlight key artefacts. In brief, the thesis deals with a description of the anatomy and properties of the retina with examples of the different stages of the aforementioned disease, an evaluation of existing solutions with a similar focus, the ideal procedure for creating an information system for use with a web application with a description of the technologies involved, and finally the method of analysing retinal images using the software functions of the system.