Manuel F M Costa
Universidade do Minho, Portugal
Title: Application of neural networks to the classification of photorefraction images
Biography
Biography: Manuel F M Costa
Abstract
Neural networks are successfully being used for many years in a large number of fields of science and technology also in medicine and virtually all fields of knowledge. Even in simple non advanced ways the results of the application of neural networks to a number of problems give good and reliable results comparable and most often better than traditional methods. In the work herein we report an application to an optometry problem: the automated classification of digital photorefraction images was obtained to characterize the refractive status of patients, mostly young children. The importance of an early evaluation of the condition of the visual system of infants is long time recognized. Non corrected optometric or ophthalmologic problems may lead to major vision and developmental non-reversible limitations in the future. Among the objective methods of refraction photorefractive techniques are specifically designed for screening young children. Over the years a number of photorefraction systems with different grades of complexity and automation were developed. One critical problem that needs to deal with in any approach of these systems is the interpretation and classification of the photorefraction images.