Identifying Natural Images From Human Brain Activity
- Technology Benefits
- Empiric assessment of visual perceptionMental control of computersMedical diagnosis
- Technology Application
- Device control and communication for people with limited physical abilitiesAssessment of attention and visual perception, such as in air traffic control and 911 operatorsEye-witness testimony verificationMedical analysis of conditions with perceptual conditions, such as migraines, epilepsy, and schizophreniaAid in diagnosis of diseases, stroke, dementia, Parkinson'sAssess the effects of therapeutic interventions (drug therapy, stem cell therapy)
- Detailed Technology Description
- None
- Supplementary Information
- Patent Number: US20130184558A1
Application Number: US13725893A
Inventor: Gallant, Jack L. | Naselaris, Thomas | Kay, Kendrick | Prenger, Ryan
Priority Date: 4 Mar 2009
Priority Number: US20130184558A1
Application Date: 21 Dec 2012
Publication Date: 18 Jul 2013
IPC Current: A61B000500 | A61B000504 | A61B00050484 | A61B000600 | A61B000603
US Class: 600409 | 600410 | 600473 | 600544
Assignee Applicant: The Regents of the University of California
Title: APPARATUS AND METHOD FOR DECODING SENSORY AND COGNITIVE INFORMATION FROM BRAIN ACTIVITY
Usefulness: APPARATUS AND METHOD FOR DECODING SENSORY AND COGNITIVE INFORMATION FROM BRAIN ACTIVITY
Summary: Method for decoding and reconstructing a subjective perceptual or cognitive experience of a human subject during imagery and recall, to identify, classify or reconstruct brain activity stimulus or mental state during psychotherapy.. Can also be used as a biofeedback device during rehabilitation, for diagnozing perceptual dysfunction due to injury and disease in neurology application, and for decoding perceptual and cognitive experiences of animals e.g. drug sniffing dogs and guide dogs.
Novelty: Method for decoding and reconstructing subjective perceptual or cognitive experience of human subject, involves reconstructing stimuli related to brain activity data based on probability of correspondence between data and activity values
- Industry
- Biomedical
- Sub Category
- Rehabilitation
- Application No.
- 9451883
- Others
-
Tech ID/UC Case
17964/2008-106-0
Related Cases
2008-106-0
- *Abstract
-
Many research inroads have been made into understanding data from human brain activity. New brain assessment devices beyond classing EEG data, such as MRIs and PET scans, have increased this available data stream. However, the information is often only inferentially related to specific brain activity.
There is an important need for individuals with limited physical capacity to control devices and communicate with others. Work towards this end has been pursued in BMI research. There is the potential in brain activity sensing devices to provide these capacities by other means as well.
Researchers at the University of California, Berkeley have made important strides in accomplishing these goals with software which can identify natural images from human brain activity. This provides an opportunity for a visual BMI. An encoding model is constructed that describes how visual stimuli are represented in the pattern of activity across visual cortex. The activity that the image produces in visual cortex has proven out to be systematically related to the particular visual stimulus that is being viewed at any point in time.
The UCB model is a variant of those that have been developed by the sensory neuroscience community over the last 50 years. The current research suggests that fMRI-based measurements of brain activity contain much more information about underlying neural processes than might have been expected. In fact so much information is available in these signals that one day it may even be possible to reconstruct the visual contents of dreams or visual imagery.
To identify which of the images elicited the measured activity the decoder scans through all possible images, and for each image it predicts what pattern of brain activity should have been elicited if that image had actually been seen. Then the decoder simply chooses the image whose predicted brain activity is most similar to the measured brain activity.
Decoding visual content is conceptually related to the neural-motor prosthesis BMI work build a decoder that can be used to drive a prosthetic arm or other device from brain activity. While the current research is focused on visual perception, other sensory systems, such as touch, taste, hearing, etc, are also amenable to analysis using the innovative software.
The potential use of this technology in the legal system brings with it most of the problems that are already known regarding eyewitness testimony.
- *IP Issue Date
- Sep 27, 2016
- *Principal Investigator
-
Name: Jack Lee Gallant
Department:
Name: Kendrick Kay
Department:
Name: Thomas Naselaris
Department:
Name: Ryan Prenger
Department:
- Country/Region
- USA
