The fundamental goal of neuroscience is to establish the link between physical events in the brain and human consciousness, from complex thoughts to emotion. Understanding the patterns of brain activity that underlie behavior is a major step toward accomplishing this goal.
Technologies to record neural activity are indispensable in our brain science endeavors because they allow us to investigate the function of unknown brain regions. We first start with a brain region of interest (let’s call it region X). Then, we make a hypothesis about its function (such as: “region x is where multiplication happens”).
From the hypothesis, we then design a behavioral task that will involve this cognitive function (“we use region x in solving math problems”). Finally, we have the subject perform the task while recording his or her activity. If the region becomes active, then we can start to make the conclusion that region X participates in the cognitive ability of multiplication.
Functional magnetic resonance imaging (fMRI) is the most well-known recording technique widely used in human cognitive neuroscience studies. Cells in the body require energy to operate, and neurons are no exception. Energy can be derived from nutrients and oxygen, which are delivered to cells around the body through the blood.
As a result, active neurons will require more energy, causing changes in the blood oxygen properties. By correlating activity levels with changes in blood, an fMRI allows us to determine which brain regions are responding to certain behavioral tasks. This technique has greatly enhanced our understanding of complex cognitive processes, including visual perception and pain.
Despite its wide utility, an fMRI has very poor temporal time scale. Neural activities can change within microseconds, while fMRI operates on the scale of seconds at best. In other words, fMRI fails to catch short bursts of activity that may actually be very important for brain function.
To catch short bursts of activity, we can turn to a technique called electroencephalography (EEG). By putting electrodes all over the scalp, we can pick up electrical activities of neurons at a much higher temporal resolution with respect to fMRI.
However, the downside here is that electrodes can only record from neurons that are toward the skull. EEGs cannot pick up activity from neurons that are located deep within the brain, such as those in the hippocampus and amygdala.
Another major disadvantage of EEG is a problem of spatial resolution. Unlike fMRI, all neuronal activities are summed up together. The end product is essentially a wave of electrical activity. This information is not useful if we are trying to figure where the activity is coming from, compounded by the fact that we are recording from many neurons at once.
In the future, I envision that we may be able to develop a recording technique that accomplishes both high temporal and spatial resolution. In other words, the ideal neuroimaging machine would be able to record activity both on a microsecond scale and single neuron level.
What I mean by single neuron level is that even an fMRI can only give information about brain activities on a macroscopic scale. For example, we can point to a region and say that it is the hippocampus, yet the hippocampus contains very intricate but tiny neural circuits composed of only a few cells. Each circuit is distinct from another, and thus we need techniques with much higher spatial resolution if we expect to dig deep.
In addition to helping us learn more about the brain, better neural recording techniques would also allow us to identify causes of brain disorders. In certain psychiatric disorders, some areas have been shown to have abnormal patterns of activity, and restoring normal activity to those sick regions could potentially reverse the disease condition.
