Method and system for correlating an image capturing device to a human user for analysis of cognitive performance
Inventors
BOTT, Nick • LANGE, Alexander R. • COSGRIFF, Rob • HSIAO, Roger • DOLIN, Brad
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-10905366-B2
Publication Date
2021-02-02
Expiration Date
Abstract
A method capturing eye movement data for detection of cognitive anomalies, includes a computer application displaying a frame on a display, capturing and displaying a video image of a user's face and eyes, while the user aligns the face to the frame, capturing and processing the face image to initiate an image eye movement capture process, outputting an indication on a display and moving the indication spatially to one of a plurality of images, capturing a video of each user eye, to track the position of the indication of the display, the image of each eye comprising a sclera portion, an iris portion, and a pupil portion, parsing the video to determine a reference images corresponding to eye positions, capturing the user's eyes while the user views familiar and novel images, and correlating the images of the user's eyes to the familiar or novel images using the reference images.
Core Innovation
A problem addressed in the background is providing a cognitive assessment that can indicate cognitive performance using eye movement data captured while presenting images to a player. The system uses a web-camera based approach to collect face/eye video while a display presents image content, and then processes the captured video data to support downstream determination of cognitive performance or diagnostic indications.
Eye tracking calibration is performed by aligning the user’s face to an on-screen frame and capturing face/eye video. A gaze model is computed using a pupil position relative to displayed dot locations, and a validation is performed to determine model usability.
During familiar and novel image phases, the system captures eye videos and derives gaze/fixation data to determine novelty preference. Novelty preference is determined from viewing time or a percentage of viewing time, and the resulting cognitive performance and/or diagnostic indications are generated after optionally filtering and validating data.
The documented architecture supports processing on a remote server or on-device, and it provides feedback and follow-up actions based on the cognitive assessment outcomes. Additional eye-related metrics such as blinks, pupil dilation, and saccades are also described as alternate or supporting measures, and the platform is described as applicable to other cognitive/behavioral assessments and downstream medical workflows.
Claims Coverage
The provided independent claim set centers on a remote matching-game display system with six inventive features. The claims focus on remote static image display, web-camera capture of eye-movement video, edited lower-resolution video data from an area of interest, host-side viewing-time and viewing-relationship determinations, success and unsuccessful indications, and related conditional outputs.
Remote static image display with selective inhibition and simultaneous presentation
Directing a remote computer system to display only static images from a first plurality on at most half of the display for a first predetermined amount of time, inhibit displaying at least one static image from the first plurality for a second predetermined amount of time, and simultaneously display a first static image from the first plurality and a second static image from a second plurality on at most half of the display for a third predetermined amount of time.
Web-camera eye-movement video capture during image display
Capturing, using a web camera of the remote computer system, video data of the player in which the video data captures eye movements while the remote computer system is displaying the first static image and the second static image.
Lower-resolution edited video data created from an area of interest
Creating edited video data from a subset of the video data in response to a pre-defined two dimensional area of interest, wherein the edited video data has a lower resolution than the video data, and providing the edited video data to the host computer.
Host-side viewing time, viewing relationship, and success flag
Determining, with the host computer, a first amount of time representing an amount of time the player views the first static image and a second amount of time representing an amount of time the player views the second static image in response to the edited video data; determining a viewing relationship between the second amount of time and the first amount of time; determining whether the viewing relationship exceeds a first threshold; and generating a success flag in response thereto.
Success indication communicated to the remote system
Providing from the host computer to the remote computer system an indication that the player is successful in response to the success flag.
Viewing relationship percentage-of-time with threshold constraint
Defining the viewing relationship for the player as a percentage of time the player views the second image, and using a first threshold greater than about 50%.
Unsuccessful indication based on absence of the success flag
Providing an indication that the player is unsuccessful in response to an absence of the success flag.
Gift delivery triggered by unsuccessful outcome
Mailing a gift to the player in response to an absence of the success flag.
Gift content as invitations to experimental drug studies
Defining the gift as invitations to participate in one or more experimental drug studies.
Mapping viewing relationship to possible physical conditions and remedial actions
Comparing the player’s viewing relationship to viewing relationships associated with possible physical conditions, determining possible conditions and corresponding remedial actions, and providing indications of the conditions and remedial actions to the remote computer system.
The claim coverage is directed to a remote matching-game display system that captures player eye-movement video via a web camera, creates lower-resolution edited video data from a pre-defined area of interest, and uses host-side timing and viewing-relationship computations to generate a success flag and communicate success or unsuccessful outcomes to the remote system, including conditional mappings to possible physical conditions and remedial actions.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?