An analysis of a model of extraocular muscle cooperation

Abstract

A model of the mechanics of human eye positioning has been evaluated for accuracy by comparison with clinical data of typical oculomotor diseases. By extending the existing monocular characterization to include interaction between a pair of eyes. a new binocular model is implemented. Furthermore. the improved iterative scheme developed for obtaining eye position from innervation of extraocular muscles insures convergence to a solution in only a few iterations. Computer programs implemented to solve this problem and its inverse are now coupled in such a way that the resulting model describes the innervational control and mechanics of positioning two eyes when one is fixating and the other is following under cover. Using this binocular model. an atlas of plots of hypothetical oculomotor defects and surgeries is compiled. Each plot simulates the prism-cover test and shows the normal left eye fixing at nine diagnostic positions and the corresponding locations of the following right eye. Abnormalities implemented for the covered eye include: a muscle palsy, a 5% weakened muscle, and a 5% strengthened muscle. In addition. surgical corrections such as myectomies, recessions, resections, and various combinations thereof are also simulated. Clinical data of 7 actual cases of oculomotor pathology are plotted in the same format as the hypothetical cases. The patient data are simulated in two different stages: First» the patient's condition both before and after surgery is simulated in accordance with a simple diagnosis. Recognizing that isolated defects rarely exist clinically, additional complications are then suggested. On the basis of hypothetical simulations in the atlas, changes in model parameters are made which yield prism-cover results more closely corresponding to the patient's pre-surgery data. Using this refined model, the simulations of surgical corrections are also improved. The results are treated in two separate cases: 1) the hypothetical cases* and 2) the clinical cases. The hypothetical results dealing with the four recti are in excellent agreement with accepted ideas. Although different from traditional views in some ways, the plots of the obliques appear reasonable. Similarly, the clinical cases show good agreement for the rectus muscles, while the obliques are not as well represented. An analysis of the results suggests that the model represents the general properties of the static extraocular system well and is a good starting place from which to develop a more accurate and sophisticated model. It is hoped that this thesis will benefit further work in this area.