The study investigated how artificial intelligence can measure changes in the posterior wall of the eye in children with early-onset high myopia. The research included 47 eyes from 31 children with an average myopia of -9.35 diopters and an axial eye length of 25.70 mm. Using a deep learning algorithm, the researchers quantified the density of the retinal mosaic in different zones and sectors of the eye. They found that this density decreased from the center of the eye towards the periphery and was higher in the nasal and inferior regions. Total background mosaic density was significantly correlated with eye length (r = 0.46), explaining 35.2% of the variability in axial length in the regression analysis. Regional mosaic density, particularly in the nasal and perifoveal regions, remained independently associated with eye elongation even after accounting for age and sex. The results suggest that AI-assisted quantification of changes in the ocular background may serve as a useful biomarker for the early diagnosis and monitoring of high myopia in children.