Data processing: measuring – calibrating – or testing – Measurement system – Statistical measurement
Reexamination Certificate
2002-03-01
2004-09-21
Barlow, John (Department: 2863)
Data processing: measuring, calibrating, or testing
Measurement system
Statistical measurement
C706S010000
Reexamination Certificate
active
06795794
ABSTRACT:
FIELD OF INVENTION
The present invention generally relates to a method for determining the probability that an event has occurred at a set of spatially localized positions in the environment, and more particularly to modeling of multisensory processing in brain maps.
BACKGROUND
All vertebrate animals constantly monitor the environment by orienting their sensory organs toward the locations of events of potential survival value. Neurobiological evidence indicates that animals utilize multisensory integration to detect the targets of orienting movements. It further indicates that the ability to integrate multisensory input is innate, and emerges as the developing brain interacts with the environment.
The superior colliculus (SC) is a major site of multisensory integration in the mammalian brain. The SC, as shown in
FIG. 1
, is located in the mammalian midbrain, and is homologous to the optic tectum of non-mammals. On grounds of differing connectivity and function, it can be divided into superficial and deep layers. The deep SC integrates multisensory input and participates in the generation of saccadic (rapid) eye movements. The superficial SC receives only visual input and does not participate in saccade generation.
The deep SC in mammals receives convergent inputs from the visual, auditory, and somatosensory systems. Sensory input arrives from many sub-cortical and extra-primary cortical regions of the brain. The deep SC sends its outputs to premotor circuits in the brainstem and spinal cord that control movements of the eyes and other structures. Neurons in the SC are organized topographically according to their receptive fields. Maps for the various sensory modalities are in register. The motor output of the SC is also topographically organized. Activation of neurons in a localized region of the SC leads, for example, to a saccade of a stereotyped direction and magnitude.
Multisensory enhancement (MSE) is a dramatic form of multisensory integration, in which the response of an SC neuron to an input of one sensory modality can be greatly increased by input of another sensory modality. MSE was first identified in the optic tectum of the rattlesnake, where visual and infrared stimuli can affect the activity of the same neurons. Percent multisensory enhancement is computed as:
%
MSE
=[(
CM−SM
max
)/
SM
max
]×100 (1)
where CM is the combined-modality response and SM
max
is the larger of the two unimodal responses. Percent MSE can range upwards of 1000%. Percent MSE is larger when the single-modality responses are smaller. This property is known as inverse effectiveness.
MSE is dependent upon the spatial and temporal relationships of the interacting stimuli. Stimuli that occur at the same time and place are likely to produce response enhancement, while stimuli that occur at different times and/or places are not likely to produce enhancement. MSE is also observed at the behavioral level. For example, a cat is much more likely to orient toward the source of a weak stimulus if it is coincident with another stimulus, even a weak one, of a different modality. MSE clearly helps animals detect targets. It is suggested that the function of MSE is to enhance the target-related activity of deep SC neurons.
Multiple observations from a variety of sensors increase the amount of information available for automated tasks such as detection and localization of events in the environment. Fusing inputs from multiple sensors involves transforming different sensor readings into a common representational format, and then combining them in such a way that the uncertainty associated with the individual sensor observations is reduced.
There are several components to the technological problem of muiltisensor fusion that have parallels with the neurobiology of the SC as described above. For example, sensor registration and alignment are issues in a multiple sensor environment. So is the implementation of a suitable, common representational format. The SC appears to solve both of these problems through the use of common topographical representations in the form of sensory maps, which allow multisensory alignment and implementation of common representational format.
SUMMARY OF THE INVENTION
The present invention relates to a method of determining spatial target probability using a model of multisensory processing by the brain. The method includes acquiring at least two inputs from a location in a desired environment where a first target is detected, and applying the inputs to a plurality of model units in a map corresponding to a plurality of locations in the environment. A posterior probability of the first target at each of the model units is approximated, and a model unit with a highest posterior probability is found. A location in the environment corresponding to the model unit with a highest posterior probability is chosen as the location of the next target.
REFERENCES:
patent: 5664021 (1997-09-01), Chu et al.
patent: 5687291 (1997-11-01), Smyth
patent: 5778082 (1998-07-01), Chu et al.
patent: 6005610 (1999-12-01), Pingali
patent: 6092058 (2000-07-01), Smyth
patent: 6115480 (2000-09-01), Washizawa
patent: 6226409 (2001-05-01), Cham et al.
patent: 6314204 (2001-11-01), Cham et al.
patent: 6552963 (2003-04-01), Baranek et al.
patent: 6567775 (2003-05-01), Maali et al.
patent: 6593956 (2003-07-01), Potts et al.
patent: 6604072 (2003-08-01), Pitman et al.
patent: 6609017 (2003-08-01), Shenoy et al.
Christopher A Seguin, Models of neurons in the superior colliculus and unsupervised learning of parameters from the series 1998, University of Illinois, Thesis, pp. 1-250.*
Anastasio, T.J., Patten, P.E., Belkacem-Boussaid, K.:Using Bayes' rule to model multisensory enhancement in the superior colliculus. Neural Computation, 12: 1165-1187. (2000).
Grossberg, S., Roberts, K., Aguilar, M., Bullock, D.:A neural model of multimodal adaptive saccadic eye movement control by superior colliculus. Journal of Neuroscience, 17:9706-9725. (1997).
Pearson, J.L., Gelfand, J.J., Sullivan, W.E., Peterson, R.M., Spence, L.D.:Neural network approach to sensor fusion. SPIE Sensor Fusion, 931: 103-108. (1988).
Rucci, M., Tononi, G., Edelman, G.M.:Registration of neural maps through value-dependent learning: modeling the alignment of auditory and visual maps in the barn owl's optic tectum. Journal of Neuroscience, 17: 334-3452. (1997).
Rucci, M., Edelman, G.M., Wray, J.:Adaptation of orienting behavior: from the barn owl to a robotic system. IEEE Transactions on Robotics and Automation, 15: 16-110. (1999).
Anastasio, T.J., Patten, P.E., Belkacem-Boussaid, K.:Using Bayes' rule to model multisensory enhancement in the superior colliculus. Neural Computation, 12: 1165-1187. (2000).
Grossberg, S., Roberts, K., Aguilar, M., Bullock, D.:A neural model of multimodal adaptive saccadic eye movement control by superior colliculus. Journal of Neuroscience, 17: 9706-9725. (1997).
Pearson, J.L., Gelfand, J.J., Sullivan, W.E., Peterson, R.M., Spence, L.D.:Neural network approach to sensor fusion. SPIE Sensor Fusion, 931: 103-108. (1988).
Rucci, M., Tononi, G., Edelman, G.M.:Registration of neural maps through value-dependent learning: modeling the alignment of auditory and visual maps in the barn owl's optic tectum. Journal of Neuroscience, 17: 334-3452. (1997).
Rucci, M., Edelman, G.M., Wray, J.:Adaptation of orienting behavior: from the barn owl to a robotic system. IEEE Transactions on Robotics and Automation, 15: 16-110. (1999).
Anastasio Thomas J.
Huang Thomas S.
Ray Sylvian R.
Barlow John
Greer Burns & Crain Ltd.
Taylor Victor J.
The Board of Trustees of the University of Illinois
LandOfFree
Method for determination of spatial target probability using... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for determination of spatial target probability using..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for determination of spatial target probability using... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3268673