Neural network analysis of the visual system functional transformation in normal aging

Abstract

Purpose. To reveal functional transformation patterns of the visual system in normal aging using neural network analysis. Material and methods. We examined 170 people aged 18 to 60 with objective refraction (under conditions of cycloplegia) from +5.5 D to –5.5 D. The criteria for selecting patients in the study groups were: maximally corrected visual acuity in the distance of each eye on a decimal scale of 1.0 and higher, normal color perception, absence of concomitant ophthalmopathology. A comprehensive assessment of the anatomical and optical eye parameters, indicators of monocular sensory reception and binocular interaction was carried out. 90 individual indicators of the visual system were used for neural network analysis with pattern recognition in a genetic algorithm with reduced dimensionality and step-by-step discriminant analysis. Results. Neural network analysis made it possible to establish the sequence of inclusion of 14 informative signs of the transformation of the visual system during normal aging. The contribution to the transformation of the visual system from changes in the accommodation system was 52%, from a decrease in the level of binocular interaction – 22%, changes in the function of the pupillary diaphragm – 13%, an increase in the temporal characteristics of sensory reception and signs of fatigue of the visual system – 11%. Conclusion. Neural network analysis allowed us to establish the sequence of inclusion of 14 informative signs of transformation of the visual system in normal aging. Normal aging of the visual system is expressed not only in a decrease in accommodative ability, but also in a decrease in the level of binocular interaction and binocular summation, in an increase in the processes of functional fatigue in the process of sensory reception, accompanied by a change in the function of the pupil. Keywords: normal aging, visual system, presbyopia, artificial neural network