DETERMINING 3D IMAGE STRUCTURES IN ADOLESCENT MYOPIA WITH DEEP LEARNING: IMPACT OF FITNESS AND GAMES

Abstract

In recent years, there has been a notable increase in myopia and refractive issues among adolescents, a trend that is concerning given the importance of visual health in overall well-being and physical activity. This study leverages Artificial Intelligence (AI) and 3D Optical Coherence Tomography Images, utilizing deep learning techniques, to analyze the refractive three-dimensional image structure of adolescent myopia. The aim is to understand how myopia and vision clarity might affect adolescents' engagement in physical fitness and games. An experimental group and a control group were established for this purpose. The analysis of the 3D image structure was conducted using sample analysis methods and fuzzy variable similarity measurement techniques. The results indicated significant variations in stereo vision and anisotropy levels among participants. In the low-level isokinetic inspection group, 34.62% had abnormal findings, while 65.38% were normal. The high anisotropy group showed 60.62% with abnormal results. Notably, in the low anisotropy group, 80.77% had normal stereo vision, but 19.23% were abnormal, with an average stereo vision of 71.92±73.92". The high anisotropy group had more acute stereo vision, with 54.55% normal and 45.45% abnormal cases, and an average stereo vision of 130.91±119.83". The findings demonstrate the effectiveness of the applied 3D image processing methods in analyzing the refractive status of adolescents. Importantly, this analysis is crucial for understanding the impact of myopia on adolescents' ability to engage in physical activities and games, which are essential for their overall physical and mental development. The study highlights the need for further research into how visual impairments affect adolescent participation in physical fitness and games, aiming to improve their quality of life and health outcomes.