Will a person learn to read thoughts?
What are you thinking about at the moment? What memories are you experiencing? Surely you think that the answers to these questions are known only to you. Neurologists are trying to look into the human mind by combining brain scans with special computer programs.
Over the past few years, researchers have learned to use patterns of brain activity to guess the images a person is looking at, his position in virtual space, as well as the decisions he intends to take. The results of recent studies demonstrate the possibility of recreating the moving images observed by the participants of the experiment – and even making assumptions about the memories they experience.
One of the most impressive achievements in this field was presented at the congress of the Society of Neurophysiologists in Chicago (Neuroscience 2009) by Jack Gallant, a leading expert in "mind reading" from the University of California at Berkeley. Together with his colleague Shinji Nishimoto, he showed that it is possible to approximately recreate the plot of a video clip viewed by a person simply by studying the activity of his brain. Scientists who attended the congress claim that such a technique of decoding nerve impulses in the future can be used for mind reading and even for the diagnosis of eating disorders.
The experiment of Nishimoto and Galant was as follows. Two laboratory employees were shown videos for two hours and simultaneously scanned their brain activity. A special computer program mapped activity patterns in the visual cortex in various aspects – shape, color, action. Then the program processed the videos posted on YouTube for 200 days, and based on the maps built earlier, it assumed a picture of brain activity that each of these videos could produce on the viewer.
Finally, the same subjects were shown a third series of video clips unknown to the computer program, and their brains were scanned. The computer compared the new tomograms with brain activity patterns suggested based on YouTube video clips. For each second of scanning, the program selected and combined 100 videos producing the closest possible nerve impulses. The result was a continuous, very blurry frame, approximately corresponding to the video clip viewed by a person.
In some cases, more successful results were obtained. When one of the participants in the experiment watched a video featuring actor Stephen Martin in a white shirt, the computer depicted a blurred silhouette of a moving man with a white torso, but the blurred spot did not resemble the moustache that Martin wears.
Despite the modest results of the first experiments, scientists express optimism and plan to improve the image reconstruction by providing the program with additional information about the content of the videos.
Videos and pictures are far from all the information that researchers are trying to extract from information about brain activity. A group of scientists from University College London, led by Eleanor Maguire and Martin Chadwick, demonstrated at a congress in Chicago that our memory is not so inaccessible to brain scanners.
A part of the brain called the hippocampus plays a major role in the formation of memory, so Maguire and colleagues focused their efforts on the study of this area. They scanned the brains of 10 people who watched three different videos in which women performed simple tasks, such as throwing away a coffee can or sending a letter. The researchers then asked the participants of the experiment to recall one of the three videos and with a probability of about 50% were able to answer which one each of the participants recalled.
According to Maguire, this figure is significantly higher than a simple coincidence, but it does not mean mind reading at all, because the program cannot decipher memories that did not pass through it. "You won't be able to connect someone to the scanner and find out what they think." The authors consider neural decoding primarily a way to understand how the hippocampus and other brain structures form memory.
Meanwhile, predicting human actions is a hot topic of research by John-Dylan Haynes from Germany. He and his colleagues from the Berlin Computer Neurology Center named after Bernstein learned how to use brain tomograms to predict the intentions of subjects in planning and performing simple tasks. Moreover, by showing images of food to people with eating disorders, scientists were able to determine which of them suffers from anorexia or bulimia based on nerve impulses from one of the "pleasure centers" of the brain.
Another aspect of neural decoding is language. Marcel Just and his colleague Tom Mitchell from Carnegie Mellon University in Pittsburgh said a year earlier that they could determine with a probability higher than random which of two words - for example, "celery" and "airplane" – the subject was thinking about. At the moment they are working on two-word phrases.
Their ultimate goal – to turn brain scans into short sentences–is, if not unattainable, then very far away. But, like the rest of the neural decoding ideas, it is as tempting as it is frightening.
Despite the successes achieved, most researchers try to avoid the words "mind reading". Emphasizing the imperfection of the method, they call it "neural decoding".
The development of mind-reading technologies also raises some concerns about the noble intentions of those who own these techniques. Without a doubt, they can be of great benefit, for example, to expand the understanding of the work of the brain or to facilitate communication with people who are unable to speak or write, like stroke victims or suffering from neurodegenerative diseases.
The Ethical RubiconBrain scanning in order to penetrate into the depths of consciousness, "read" thoughts and memories is a very tempting and daring idea.
However, neural decoding techniques still remain limited in their application. At the moment, they only work if the brain has already been repeatedly scanned and under very specific experimental conditions. Can these attempts be called mind reading? And is it worth worrying about the frightening possibilities of such technologies?
To some extent, this is a matter of interpretation of the phrase "mind reading", and yet many experts in the field of neurophysiology argue that the results achieved to date are far from what people understand by the expression "mind reading" – for example, to determine whether a terrorist is going to detonate a bomb on an airplane.
Despite the very vague prospects for the use of such technologies, it is already time to think about the ethical side of this issue, says John-Dylan Haines.
Some companies claim that brain scans will help identify lies or determine whether a commercial is valid, and there is some truth in these statements. According to Haines, it is necessary to determine what neural decoding can and cannot reliably do in order not to undermine public confidence in this field of research.
Neural decoding can be a double-edged sword. If hopes for this technology are ever realized, the same machine that reads the thoughts of patients with neurodegenerative diseases can become a dangerous tool in dirty hands.
Ruslan Kushnir
Portal "Eternal youth" http://vechnayamolodost.ru based on New Scientist: Brain scanners can tell what you're thinking about05.11.2009