MRI imaging of eye shape and associations with myopia in Singapore adults

Abstract

Purpose : To compare eye shape across the refractive error range using brain magnetic resonance imaging (MRI) images in Singapore adults.

Methods : A total of 450 participants aged 50 years and above (64.2 +/- 6.5 years old) in the Singapore Epidemiology of Eye Disease (SEED) study (169 Chinese, 165 Indians and 116 Malays) were included. Refraction was performed using an autorefractor, and axial length (AL) was measured using optical biometry (IOLMaster, Zeiss). The MRI volumes (3T Magnetom Trio Tim; Siemens) of the right eye were semi-automatically segmented in the axial planes, and posterior eye radius of curvature (Rc) (related to total eye volume) and asphericity (Qxz with > 0 as oblate, < 0 as prolate) were calculated. Multivariate liner regression models were used to identify Rc and Qxz and associations with spherical equivalent (SE) and AL, controlling for confounders.

Results : Among 441 eligible subjects, the prevalence of myopia (-0.50 Diopter (D) > SE) and high myopia (-5.00D > SE) were 25.6% and 3.6%, respectively. Rc was correlated positively with AL (Pearson correlation coefficient, r = 0.56) and negatively with SE (r = -0.27). Qxz was correlated negatively with AL (r=-0.50) and positively with SE (r=0.44). Both larger eye volume and prolate eye shape were correlated with longer AL and worse SE. Multivariable regression analysis revealed that larger eye volume, (larger Rc) was associated with worse SE (B = -1.8, 95% confidence interval (CI) -2.3 to -1.3, p < 0.001)), longer AL (B =1.4, CI 1.2 to 1.6, p < 0.001) and myopia (odds ratio (OR) = 4.3, CI 2.6 to 7.42, p < 0.001). Lower Qxz indicating increasing axial prolateness was also associated with worse SE (B = 6.6, CI 5.6 to 7.9, p < 0.001), longer AL (B= -3.5, CI -4.0 to -3.0, p < 0.001) and myopia (p<0.001). Comparison among 3 groups with different SE: Non-myopia (SE > -0.50 D), mild-moderate myopia (-0.50 D > SE > -5.00 D) and high myopia (SE < -5.00 D) showed only 10% of mild-moderate myopic eyes were prolate, >50% of high myopic eyes were prolate.

Conclusions : We have demonstrated that eye volume and the prolate eye shape were associated with myopic eyes, and that the prolate changes of myopic eyes become more prominent in the group with high myopia. Clinically, our findings are important because eye shape may be an important imaging biomarker for high myopia and pathologic myopia changes.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.