3 A 3-DIMENSIONAL MEASUREMENT METHOD FOR EYE MOVEMENT AND FULLY AUTOMATED DEEP-LEARNING BASED SYSTEM FOR VERTIGO DIAGNOSIS

摘要

The present invention provides a method and a system for measuring and diagnosis eyeball movement, which process an image obtained from an image sensor, measure movement of an eyeball in a three-axis, determine possible diagnosis based on the movement of the eyeball and provide the diagnosis to a user. The eyeball movement measuring method comprises: a step (a) of presenting a standardized nystagmus test item, and receiving the image of the eyeball according to the nystagmus test item; a step (b) of recognizing a pupil and an iris from the image of the eyeball; a step (c) of calculating a horizontal and vertical change amount of the pupil and a rotational momentum of the iris; a step (d) of judging a change value and a directivity for a three-axis direction of the eyeball based on the horizontal and vertical change amount of the pupil and the rotational momentum of the iris; and a step (e) of generating a diagnosed result for vertigo through a deep-learning modeling based on the change value and the directivity for the three-axis direction of the eyeball.