New scientific tool decodes brain signals intended for body’s speech-related movements and turns them into computerized speech, as reported by Axios. While the device has limitations, it’s a step forward in helping those who have lost the ability to speak. However, neural imaging is often misunderstood.
Currently, many speech translators for non-verbal people work painfully slow while the user types, clicks, or moves their eyes to find and input one word at a time. According to Axios, the new technology in development “decodes brain signals for the speech-related movements of the jaw, larynx, lips, and tongue” and turns them into computerized speech. The device is still in its early stages of development and its scope is limited, but it has great—and ominous-sounding—potential. Let’s separate neural imaging facts from fiction.
Translating Thought into Speech – Popular Methods of Measuring Brain Activity
One of the most common ways to measure neural activity is with an electroencephalogram, or EEG. Using an EEG, “We can record electrical potentials—that is, averaged neural activity across many, many neurons—by placing electrodes on the scalp,” said Dr. Indre Viskontas, Professor of Psychology at the University of San Francisco. Dr. Viskontas, who is also Professor of Sciences and Humanities at the San Francisco Conservatory of music, said that the EEG gives neurologists an idea of which groups of neurons are doing what at very precise moments.
What about cell activity as a signal device? “We can also see where cells have just been active by tracking the amount of oxygen in the blood flowing to that region,” Dr. Viskontas said. “Since active cells require oxygen, we can use magnetism to detect changes in the oxygenation of blood in different parts of the brain.” This method of measuring the activity of the brain is the MRI and fMRI—magnetic resonance imaging and functional magnetic resonance imaging, respectively. Dr. Viskontas explained that the MRI came first and its tools were modified and adapted to measure activity in the brain, which is now known as the fMRI.
The third common way to detect brain activity is actually on its way out. “PET scans, or positron emission tomography, are also still used—though less frequently, because they provide much of the same information that we can garner from fMRI, but involve the injection of a radioactive substance, which is then tracked as it’s taken up by active cells,” Dr. Viskontas said. “It’s largely reserved these days to diagnose neurodegenerative diseases or cancer, rather than to investigate cognition.”
Common Misconceptions of Neural Imaging
Neural imaging has raised many concerns about mind-reading, brain damage, and more—most of which are unfounded. However, these misconceptions often stem from improper understanding of parts of the brain—especially the amygdala and the reward circuitry of the brain.
“The media often misinterprets amygdala activation as a sign that we’re fearful of something or that our behavior is somehow driven by fear,” Dr. Viskontas said. “But what if the emotions we feel during [an] event are positive? It turns out that the amygdala notices those, too, and plays a role in making sure that you learn what it is that lead to a positive result.” The size and condition of the amygdala affects sexual, aggressive, and maternal behaviors in animals as well as depression, anxiety, PTSD, and borderline personality disorder in humans.
Dr. Viskontas also explained that the reward circuitry of the brain is involved in the release of dopamine, a neurotransmitter associated with both pain and pleasure. “We can’t infer that someone is feeling pleasure just because their reward pathways light up,” she said. “The reverse inference doesn’t necessarily hold.”
Before we make conjectures about brain scans, thought invasion, and other sci-fi elements that come from the study of neurology, we should learn to understand the tools used for measuring brain activity and the characteristics of the parts of the brain involved. The new speech translation device is a notable step forward in neural imaging, but we’re still far from developing 100-percent accurate lie detectors, much less full digital uploads of the brain system.
Dr. Indre Viskontas contributed to this article. Dr. Viskontas is an Adjunct Professor of Psychology at the University of San Francisco and Professor of Sciences and Humanities at the San Francisco Conservatory of Music, where she is pioneering the application of neuroscience to musical training.