Macular translocation surgery: computer simulation of visual perception.

Abstract

BACKGROUND: Macular translocation can be associated with visual improvement, but patients often experience symptoms of confusion or diplopia. There is a high incidence of suppression of the operated or the fellow eye. The aim of this study is to use computer software to examine the pre- and post-operative fundal images, in order to better understand how patients see after macular translocation surgery. METHODS: We created a graphical user interface that allowed a user to identify and record common landmark points in pre- and post-operative fundal images. We used these points to carry out interpolations using two algorithms, namely bilinear and thin-plate spline transformations. The transformations were applied to the Mona Lisa in order to appreciate how patients might see. RESULTS: Given two sets of corresponding points, both algorithms were able to approximate the effect of the surgery. Bilinear transformation was able to account for changes to the retina as a whole, including rotation, stretches, compression and shear. The thin-plate spline algorithm additionally accounted for the considerable regional and uneven local effects. Applying the later algorithm to the Mona Lisa produced inconsistent and warped images. CONCLUSIONS: Our results confirmed that neurosensory redistribution was associated with most cases of MT360. We infer from these results that corresponding retinal elements between two eyes would no longer correspond after surgery. The distortion of images from the operated eye could not be completely corrected by squint surgery, and this may account for the high incidence of suppression of the fellow or the operated eye after surgery.