US 2020/0345288 A1 · A 3-DIMENSIONAL MEASUREMENT METHOD FOR EYE MOVEMENT AND FULLY AUTOMATED DEEP-LEARNING BASED SYSTEM FOR VERTIGO DIAGNOSIS

Abstract

The present invention relates to an eye movement measurement method including the steps of: (a) presenting standardized nystagmus test items and receiving am image of an eye in accordance with the standardized nystagmus test items; (b) recognizing a pupil and an iris from the image of the eye; (c) calculating amounts of horizontal and vertical changes of the pupil and an amount of torsional movement of the iris; (d) determining change values and orientations for three axis directions of the eye on the basis of the amounts of the horizontal and vertical changes of the pupil and the amount of the torsional movement of the iris; and (e) generating a diagnosis result for vertigo through deep learning modeling on the basis of the change values and orientations of the three axis directions of the eye.

Claims17

Dependent claims

An eye movement measurement and diagnostic method comprising the steps of:

(a) presenting test items and receiving an image of an eye, which is acquired for each of the test items, for each test;
(b) recognizing a pupil and an iris from the image of the eye;
(c) calculating amounts of horizontal and vertical changes of the pupil and an amount of torsional movement of the iris;
(d) determining change values and orientations for three axis directions of the eye on the basis of the amounts of the horizontal and vertical changes of the pupil and the amount of the torsional movement of the iris;
and
(e) generating a diagnosis result for vertigo on the basis of the change values and orientations of the three axis directions of the eye.

3 more elements

An eye movement measurement and diagnosis system comprising:

an image processing apparatus;
and
a diagnosis apparatus,
wherein the image processing apparatus comprises:
an image receiver configured to receive an image of an eye acquired for each test item;
a pupil recognizer configured to recognize a pupil from the image of the eye;
an iris recognizer configured to recognize an iris from the image of the eye;
a change amount calculator configured to calculate amounts of horizontal and vertical changes of the pupil and an amount of rotational movement of the iris;
and
a change amount processor configured to determine change values and orientations of three axis directions of the eye on the basis of the amounts of the horizontal and vertical changes of the pupil and the amount of the torsional movement of the iris, and
wherein the diagnosis apparatus generates a diagnosis result for vertigo on the basis of the change values and orientations of the three axis directions of the eye received from the image processing apparatus.

8 more elements

Dependencies are inferred from the claim text.

A 3-dimensional measurement method for eye movement and fully automated deep-learning based system for vertigo diagnosis

Life of this patent

A 3-dimensional measurement method for eye movement and fully automated deep-learning based system for vertigo diagnosis

Life of this patent

Abstract

Claims17