Eye-tracking method and system for screening human diseases

• US 8,808,195 B2
• Filed: 01/15/2010
• Issued: 08/19/2014
• Est. Priority Date: 01/15/2009
• Status: Active Grant

• Current Assignee: Queen's University At Kingston, University of Southern California
• Original Assignee: Queen's University At Kingston
• Inventors: Tseng, Po-He, Cameron, Ian G. M., Munoz, Douglas P., Itti, Laurent
• Primary Examiner(s): Dougherty, Sean
• Assistant Examiner(s): Stout, Michael C
• Application Number:
  US12/656,109
• Publication Number:
  US 20100208205A1
• Time in Patent Office: 1,677 Days
• Field of Search: 600/558, 600/595
• US Class Current: 600/558
• CPC Class Codes: A61B 3/113 for determining or recording eye movement
• CPC Class Codes: G06V 10/454 Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]
• CPC Class Codes: G06V 30/274 Syntactic or semantic context, e.g. balancing
• CPC Class Codes: G06V 40/19 Sensors therefor

Abstract

This invention provides methods, system, and apparatus for assessing and/or diagnosing a neurobehavioural disorder in a subject. The methods, systems, and apparatus include the subject freely observing a visual scene, without having to carry out a task or follow specific instructions. In one embodiment, a computational model is used to select one or more feature in a visual scene and generate a spatial map having first map values that are predictive of eye movement end points of a hypothetical observer relative to the one or more feature. A subject'"'"'s eye movements are recorded while the subject freely observes the visual scene, and a difference between second map values that correspond to the subject'"'"'s eye movement endpoints and a set of map values selected randomly from the first map values is quantified, wherein the difference is indicative of a neurobehavioural disorder in the subject. Neurobehavioural disorders such as Parkinson'"'"'s disease, Alzheimer'"'"'s disease, Huntington'"'"'s disease, fetal alcohol spectrum disorder, attention deficit hyperactivity disorder, schizophrenia, autism, Tourette syndrome, and progressive supranuclear palsy may be assessed and/or diagnosed.

Claims

1. A method of assessing, diagnosing, or assessing and diagnosing a neurobehavioural disorder in a subject, comprising:

using a processor-based computational model to select one or more feature in a visual scene and generate a spatial map having first map values that are predictive of eye movement end points of a hypothetical observer relative to the one or more feature;

recording the subject'"'"'s eye movements while the subject freely observes the visual scene; and

generating second map values that correspond to the subject'"'"'s eye movement endpoints that do not rely on the subject performing a task;

quantifying a difference between the second map values and a set of map values selected randomly from the first map values;

wherein the difference is indicative of a neurobehavioural disorder in the subject.

2. The method of claim 1, wherein the difference is indicative of the extent to which the subject was guided towards the selected one or more feature while observing the visual scene.

3. The method of claim 1, wherein the one or more feature is a saliency feature, the spatial map is a saliency map, and the map values are saliency values.

4. The method of claim 3, wherein the difference is indicative of the extent to which the subject was guided towards one or more saliency feature while observing the visual scene.

5. The method of claim 1, wherein the visual scene includes a video displayed on a television or computer screen.

6. The method of claim 5, wherein a spatial map is generated for each frame of the video, and one or more eye movements are recorded for each frame of the video.

7. The method of claim 6, wherein the video comprises a series of video clips.

8. The method of claim 1, wherein the computational model selects one or more feature from the group consisting of a statistical analysis of the gist of the visual scene, detection of human face(s), detection of person(s), detection of animal(s), detection of landscape(s), and detection of text message(s).

9. The method of claim 1, wherein the neurobehavioural disorder is selected from Parkinson'"'"'s disease, Alzheimer'"'"'s disease, Huntington'"'"'s disease, fetal alcohol spectrum disorder, attention deficit hyperactivity disorder, schizophrenia, autism, Tourette syndrome, and progressive supranuclear palsy.

10. Apparatus for assessing, diagnosing, or assessing and diagnosing a neurobehavioural disorder in a subject, comprising:

a visual scene and a spatial map corresponding to the visual scene, the spatial map having first map values that are predictive of eye movement end points of a hypothetical observer observing the visual scene, relative to the one or more feature;

an input for receiving data corresponding to eye movements of a subject observing the visual scene, the data including second map values that correspond to the subject'"'"'s eye movement endpoints that do not rely on the subject performing a task; and

a processor that quantifies a difference between the second map values and a set of map values selected randomly from the first map values, and outputs the difference;

wherein the difference is indicative of a neurobehavioural disorder in the subject.

11. The apparatus of claim 10, wherein the visual scene includes a video.

12. The apparatus of claim 11, wherein a spatial map is provided for each frame of the video.

13. The apparatus of claim 11, wherein the video comprises a series of video clips.

14. The apparatus of claim 10, including a video display that displays the visual scene.

15. The apparatus of claim 10, including an eye tracker that records eye movements of the subject observing the visual scene, or outputs to the processor data corresponding to eye movements of the subject observing the visual scene.

16. The apparatus of claim 10, wherein the neurobehavioural disorder is selected from Parkinson'"'"'s disease, Alzheimer'"'"'s disease, Huntington'"'"'s disease, fetal alcohol spectrum disorder, attention deficit hyperactivity disorder, schizophrenia, autism, Tourette syndrome, and progressive supranuclear palsy.

17. The apparatus of claim 10, further comprising a computer program stored on non-transitory storage media compatible with the processor, the computer program containing instructions to direct the processor to use a computational model to select one or more feature in the visual scene and generate the spatial map having first map values that are predictive of eye movement end points of a hypothetical observer relative to the one or more feature.

18. A method of assessing, diagnosing, or assessing and diagnosing a neurobehavioural disorder in a subject, comprising:

monitoring the subject'"'"'s eye movements while the subject freely observes a visual scene and preparing second map values that correspond to the subject'"'"'s eye movement endpoints, wherein the second map values correspond to the subject'"'"'s eye movement endpoints that do not rely on the subject performing a task;

quantifying a difference between the second map values and a set of first map values;

wherein the first map values were obtained using a processor-based computational model that selects one or more feature in the visual scene and generates spatial map values comprising the first map values, the spatial map values being predictive of eye movement end points of a hypothetical observer relative to the one or more feature;

wherein the difference is indicative of a neurobehavioural disorder in the subject.

19. The method of claim 18, wherein the set of first map values are randomly selected from the spatial map values.

20. The method of claim 18, wherein the difference is indicative of the extent to which the subject was guided towards the selected one or more feature while observing the visual scene.

21. The method of claim 18, wherein the one or more feature is a saliency feature, the spatial map is a saliency map, and the map values are saliency values.

22. The method of claim 21, wherein the difference is indicative of the extent to which the subject was guided towards one or more saliency feature while observing the visual scene.

23. The method of claim 18, wherein the visual scene includes a video displayed on a television or computer screen, or a series of video clips displayed on a television or computer screen.

24. The method of claim 23, wherein a spatial map is generated for each frame of the video, and one or more eye movements are recorded for each frame of the video.

25. The method of claim 18, wherein the neurobehavioural disorder is selected from Parkinson'"'"'s disease, Alzheimer'"'"'s disease, Huntington'"'"'s disease, fetal alcohol spectrum disorder, attention deficit hyperactivity disorder, schizophrenia, autism, Tourette syndrome, and progressive supranuclear palsy.