A group of cognitive neuroscientists and computer scientists in the Carnegie Mellon University have successfully developed a technique to tell what other people are thinking, by observing their brain activity.
When a person thinks of some particular object, thoughts and perceptions are generated in various positions of the brain producing a particular pattern specific for that object. Different objects cause different patterns. By combining medical imaging and machine learning methods, scientists have been able to develop a novel algorithm—a set of well defined instructions to classifying them according to their corresponding objects. As a result, by analyzing those patterns, one can exactly tell what others were thinking about in a particular moment.
In addition, when this algorithm is trained from datasets obtained from a particular person, it could correctly classify unknown datasets obtained from a different person showing that there are similarities between brain activities of two different persons. "This part of the study establishes, as never before, that there is a commonality in how different people’s brains represent the same object," says Mitchell, head of the Machine Learning Department in Carnegie Mellon's School of Computer Science.
Finally researchers are hoping to develop this new technique further in order to use this as a valuable tool for studies and research in neurological disorders, such as autism.
When a person thinks of some particular object, thoughts and perceptions are generated in various positions of the brain producing a particular pattern specific for that object. Different objects cause different patterns. By combining medical imaging and machine learning methods, scientists have been able to develop a novel algorithm—a set of well defined instructions to classifying them according to their corresponding objects. As a result, by analyzing those patterns, one can exactly tell what others were thinking about in a particular moment.
In addition, when this algorithm is trained from datasets obtained from a particular person, it could correctly classify unknown datasets obtained from a different person showing that there are similarities between brain activities of two different persons. "This part of the study establishes, as never before, that there is a commonality in how different people’s brains represent the same object," says Mitchell, head of the Machine Learning Department in Carnegie Mellon's School of Computer Science.
Finally researchers are hoping to develop this new technique further in order to use this as a valuable tool for studies and research in neurological disorders, such as autism.
(same article on http://compuzer.blogspot.com/2008/03/brain-reader.html by same author)
Read the original article ; Category : Computer Science, Neurobiology
