Ask An Engineer

How are thoughts measured?

By watching one neuron at a time, or by looking at how millions of them are talking to each other…

By Elizabeth Dougherty

“We’re limited in our ability to record individual human neurons,” says Charles Jennings, director of neurotechnology at the MIT McGovern Institute for Brain Research. He points, however, to one major exception. It is possible during neurosurgery, when the brain is exposed and the patient is awake, to measure the activities of individual nerve cells in the brain. Doctors can show the patient something—a picture or words—and measure the response to the stimulus.

This kind of experimentation has led to the discovery that single neurons may show extraordinarily precise patterns of activity. For instance, in one study a single neuron lit up in response to actress Halle Berry—regardless of whether the patient saw her name, her face as she walked the red carpet, or Berry in the role of Catwoman.

For the most part, however, scientists measure human brain activity en masse. They do it non-invasively using tools such as functional magnetic resonance imaging (fMRI). “These tools aren’t sensitive enough to record the activity of a single neuron” says Jennings. “It’s more like an airplane view of traffic. It’s hard to see an individual car, but it’s easy to see if it’s rush hour.” A functional MRI detects brain activity by measuring changes in blood flow and blood oxygen levels. As neurons fire, they use up oxygen in the blood. Consequently, the magnetic signature of the blood’s hemoglobin, which carries oxygen, also changes. The MRI picks up this change and uses it to reveal thought “hot-spots.”

Recently, researchers at MIT developed a new kind of MRI sensor that could reveal brain activity more directly. Developed by Alan Jasanoff, a professor of Biological Engineering and an associate member of the McGovern Institute, the sensor measures concentrations of dopamine, a brain chemical that carries messages between neurons and is involved in learning, movement, and other activities. Like a standard MRI, it picks up a magnetic change in a protein—only in this system, the magnetic signature changes when the protein binds to dopamine rather than oxygen.

This new sensor comes with some practical challenges: the magnetic protein it relies on is biologically engineered and must be directly injected into the brain to work, and it only detects dopamine in areas local to the injection. The team is looking into easier ways to get the protein into the brain, such as using chemotherapy delivery methods to get drugs past the blood-brain barrier, or genetically engineering brain cells so they can make the protein themselves.

Jasanoff’s lab is also developing sensors that detect other neurotransmitters besides dopamine. “By actually detecting a specific molecular component of what the brain is doing,” says Jasanoff, “we can say something much more specific about its activity and ultimately about its circuitry and mechanisms than we could if we used a conventional blood-related fMRI.”

Thanks to Shashank Singh for this question.

Submit A Question

content Link link