Everything You Need to Know about Noninvasive Functional Brain Imaging
By Sheeva Azma
You have probably seen reporting of brain imaging studies in the news. These studies typically feature a “fancy” image of the brain which has several different blobs, sometimes in many different colors, on an MRI scan of the brain. Typically, these studies, which are often reported as popular science articles in the news media, claim to have discovered a new area of the brain responsible for various brain functions. But what exactly is brain imaging, and what do these studies actually mean? Can a simple brain scan discover the mental and neuronal origins of human consciousness?
In this blog post, Sheeva Azma of Fancy Comma, LLC explains various brain imaging methods. Sheeva has a Bachelor’s degree in Brain & Cognitive Sciences from MIT, as well as a Master’s degree in Neuroscience from Georgetown, as well as over 10 years working in the field of cognitive neuroscience. Check out Sheeva’s brain imaging research contributions here.
What Is Functional Neuroimaging?
Functional neuroimaging refers to the use of brain imaging methods to be able to look into the activities of the brain. Functional imaging is typically non-invasive in nature, though earlier types of this type of brain imaging did include methods that were slightly more invasive. For example, a type of brain imaging called Positron Emission Tomography or PET scanning uses radioactive tracers.
Functional neuroimaging exists in contrast to structural neuroimaging, which is brain imaging used to probe the structure of the brain. Examples of structural neuroimaging include Computerized Tomography (often called a CT or CAT scan) and structural MRI scans.
Why Is Functional Neuroimaging Important?
Functional brain imaging is used to look at how your brain works. That’s extremely important if we want to understand the neural underpinnings of diseases and disorders, or everyday aspects of our brain function — for example, language or something like decision-making or behavioral problems.
Functional neuroimaging is extremely versatile and can be used in both hospital and research settings. Scientists can use these brain imaging methods to see how the brain works. These methods are also used by clinicians — especially neurologists — to offer a glimpse into patients’ brain function or to help patients prepare for brain surgery. As functional neuroimaging reveals more interesting and useful insights about how the brain works, it will become an even more useful tool.
The purpose of this post is to provide a quick primer on the different types of functional neuroimaging methods that exist today. You should know that, in reality, there are many many different types of methods. That’s why this post aims to elucidate the main concepts.
Noninvasive vs. Invasive Brain Imaging
We’ve already talked about the difference between functional and structural imaging. Beyond that, there are other dichotomies in the world of brain imaging. For example, there are two main kinds of brain imaging techniques: invasive and non-invasive.
Non-invasive techniques don’t require you to undergo surgery. In fact, non-invasive techniques do not introduce anything into the body.
Invasive techniques, on the other hand, may involve placing electrodes directly into your brain or being injected with a radioactive tracer. These techniques carry higher risks, so they are only used when absolutely necessary in a hospital setting.
Examples of invasive techniques include intracranial electroencephalography or iEEG, which is used to measure and monitor the brain activity of epileptic patients going in for surgery.
Another example of an invasive technique is Positron Emission Tomography or PET scanning. PET imaging involves injection of a radioactive tracer; the tracer binds to metabolic compounds in the bloodstream, such as oxygen and/or glucose, to help clinicians and researchers look at patterns of metabolic activity in the brain.
In reality, performing iEEG or PET imaging in the larger population for research purposes would be unethical because any benefit to the study participant would be much less than the risks.
The good news is that there is another type of imaging that carries less risks. This type of imaging is called noninvasive brain imaging because it does not introduce anything new to the body or require surgery. This is why most people can easily go to their doctor to get an MRI or EEG over the course of a few hours and have no ill effects. You should also know, though, that noninvasive imaging’s inability to directly access the brain (such as with iEEG or a radioactive tracer) does mean that there are some limitations to this important and widely used technology.
What Makes Non-Invasive Brain Imaging So Important?
One thing that makes non-invasive functional imaging special is that it allows us to look into human brain function. Due to the high level of risk involved, it is not ethical to conduct invasive brain research in an invasive fashion on humans. That means that non-invasive brain imaging methods are one of the only ways by which we can understand what makes humans human.
Since more and more non-invasive functional brain imaging methods are in use, the purpose of this article is to give you a quick overview of these methods. Especially with the rise of artificial intelligence (AI) and machine learning, the data being generated by non-invasive functional imaging is paving the way for literal mind-reading. In other imaging realms, AI can even outperform pathologists looking for prostate cancers on pathological screening. Computers can now decode patterns of human brain activity. This makes AI-powered brain imaging analysis very exciting.
Keep reading to learn more about the three main types of noninvasive functional neuroimaging: electroencephalography (EEG), magnetoencephalography (MEG), and functional magnetic resonance imaging (fMRI). Each of these measures different things — EEG measures electrical fields, while MEG measures magnetic fields, and fMRI creates maps of what’s known as “brain activation.”
To get an idea of noninvasive functional brain imaging and its historical context, let’s first discuss the timeline of development for these three different brain imaging modalities — EEG, MEG, and fMRI — and provide a brief overview of what they measure.
Timeline of Noninvasive Functional Neuroimaging
- Electroencephalography, or EEG, is the oldest of the three methods. Noninvasive EEG involves recording electrical potentials from the surface of the scalp — as we’ve mentioned, there’s also an invasive form of EEG, called iEEG, that is used in epilepsy patients going in for surgery. EEG was invented in 1929 by a German doctor named Hans Berger.
- Magnetoencephalography, or MEG, is a method similar to EEG. MEG noninvasively measures magnetic fields arising from the brain’s cells. MEG was invented at the Massachusetts Institute of Technology by David Cohen in the early 1970s.
- There are now many types of functional magnetic resonance imaging, or fMRI, but the first type of fMRI to be discovered was BOLD fMRI in 1990 by Japanese researcher Seiji Ogawa. BOLD fMRI measures brain activation by way of looking at the flow of oxygenated blood in the brain. The exact relationship BOLD fMRI and brain cell activity remains unclear.
Electroencephalography or EEG works by measuring the electrical currents on the surface of your scalp. The idea here is that your brain cells, or neurons, communicate via electrical messages. These electrical messages generate electrical fields which can be detected on the surface of your scalp. That’s not to say that you’re walking around generating electricity left and right. The electrical fields that your brain cells are generating are, in reality, minute — on the order of microvolts (mV), or 10-6 volts.
A typical EEG setup involves placing electrodes on your scalp in specific places to then measure brain activity. Though EEG is the oldest of the three noninvasive brain imaging methods surveyed here, the typical EEG setup can look very futuristic:
As we’ve discussed, EEG is a non-invasive imaging method. Recall that non-invasive means that the imaging method does not require any type of surgery or procedure. Instead, you simply put the electrodes on your scalp (well, the clinician or researcher would do that part), and the measurement of brain activity is good to go. Of course, the signals generated by your brain are very complex, so software and computers are needed to decode exactly what they mean.
One type of EEG study is called Event-Related Potential (ERP). ERP studies seek to identify certain waveforms associated with different types of brain activity. These waveforms can be positive or negative, and are usually classified by their latency — or the time after a stimulus. For example, the N400 waveform can be observed as a wave with negative amplitude (hence the N) that is occurring 400 milliseconds after you hear or read a word. The N400 is a semantic marker of incongruity. If you hear a sentence like “I like my coffee with nails,” your brain will create an N400 waveform because the sentence was supposed to be something like “I like my coffee with cream.” The unexpected nature of the word “nails” in that sentence causes your brain to produce the specific N400 waveform. Certain patient populations, such as schizophrenics, have a lower N400 response than healthy people.
EEG is one of the most versatile imaging methods currently available. It’s inexpensive, easy to use, and it lends itself to a variety of different applications. One major disadvantage of EEG is its bad spatial resolution, which makes it hard to figure out where the electrical potentials are coming from in your brain. On the other hand, the time resolution of EEG is excellent. This means that brain activity can be quickly identified, but figuring out where the brain activity is coming from is more complex.
A Quick Note about Intracranial Electroencephalography (iEEG)
There’s, in reality, a more invasive kind of EEG called intracranial electroencephalography. This type of EEG is also called iEEG or intracranial EEG. Intracranial EEG is considered the gold standard for understanding brain activity. That’s because, for an intracranial EEG, electrodes are placed in specific areas of the brain. This ensures that you can localize and record activity from specific neurons, making it possible to get very detailed information about brain activity from an iEEG. The disadvantage here is that because it is so invasive, the iEEG is rarely used. For the most part, an iEEG is never used except in epilepsy surgery. This is because in epilepsy surgeries, a small part of the brain is often ablated, or taken out, and functional imaging such as an iEEG is needed to map the function and structure of the brain to make sure that the surgery only takes out the part of the brain that it needs to. In the real world, people do not regularly go in for an iEEG.
Another type of imaging modality is called magnetoencephalography, or MEG.
Due to some complicated mathematics which we won’t get in to here, MEG is thought to measure the magnetic fields being emitted by the neurons (that is, brain cells) in the cortex of the brain’s cerebral cortex — the topmost layer of brain cells. The cortex is the seat of higher-order cognitive functions, sensory processing, motor processing, decision-making and other faculties.
Magnetoencephalography requires very small sensors called SQUIDs (Superconducting Quantum Interference Devices) that can measure magnetic fields on a femtotesla (or fT, 10–16 tesla) scale. There are minute differences between MEG and EEG, but for the purposes of this article, we will not discuss them.
Both the MEG and EEG benefit from excellent temporal resolution. That means that it can understand brain activity on the millisecond level. However, the problem with both MEG and EEG is that they make it difficult to resolve where the signal is coming from in the brain. Spatial resolution is where functional MRI shines — we’ll talk about that in the next section.
Functional Magnetic Resonance Imaging (fMRI)
Functional magnetic resonance imaging, also called functional MRI or fMRI, is a type of imaging which uses the principles of MRI to measure the activities of the brain on a much slower timeframe than MEG or EEG. While MEG and EEG both have millisecond resolution, fMRI can have a temporal resolution on the order of 2 to 10 seconds.
We’ve all seen fMRI images in the news. Typically these images consist of an X-ray like image of the brain with a blob of color somewhere on the brain. This blob supposedly represents a part of our brain that is activated whenever we do a certain task.
Different Types of fMRI
There are many types of functional MRI. The most common type of fMRI is called BOLD (Blood Oxygenation Level Dependent) fMRI. BOLD fMRI looks at the hemodynamic response of the brain. In other words, it looks at the blood flow of the brain and where blood, specifically oxygenated blood, travels to help the brain conduct its needed functions. Because the brain needs oxygen, looking at the places where oxygen molecules travel in the brain can give you a sense of different areas of the brain are being used.
There are many other types of functional imaging beyond BOLD that study other aspects of brain function. A measure called diffusion tensor imaging or DTI allows you to look at white matter tracts in the brain. White matter tracts reflect the structural integrity of white matter, which is the insulation covering the “wiring” of the brain.
Another type of functional imaging is called Arterial Spin Labeling (ASL). Arterial spin labeling looks at blood flow in the brain more generally than BOLD fMRI, which focuses primarily on blood flow relative to oxygenation level.
Disadvantages of fMRI
fMRI requires very specific equipment — namely, an expensive MRI scanner. Access to an MRI is not always possible, so that is one drawback of functional magnetic resonance imaging. The other drawback is that the functional MRI image acquisition is on a very slow time scale of about two or more seconds — so it’s very difficult to resolve the timeframe of activation. However, the spatial resolution of fMRI is excellent.
The other problem that exists relating to BOLD fMRI is that it’s unclear exactly how brain activation causally relates to the activity of individual neurons or even groups of neurons. What does it mean when neurons need more oxygen? This is the question that is still being worked on and trying to be understood by scientists.
Challenges of Functional Neuroimaging
While non-invasive functional imaging techniques seem complex, they are important to learn about, as they’re actually very powerful means to learn about how the brain works.
Sadly, there are a lot of theoretical and real-world limitations associated with these different brain imaging techniques. Many of them are too complex to discuss in a short blog post. I’ve summarized some of the main challenges here.
Each Method Has Technological Limitations
Each imaging technique comes with its own advantages and disadvantages. That’s why different imaging methods are sometimes used simultaneously — such as MEG + structural MRI — in order to try to model a specific cognitive process. The use of EEG or MEG and structural methods together can help overcome the spatial resolution issues inherent in these different techniques. In recent decades, researchers have also sought to combine these three methods — fMRI, MEG, and EEG — in order to gain more precise insights regarding the timeframe of fMRI activations, with the goal of figuring out what is really happening in fMRI on the neural level.
The Statistics of Neuroimaging Can Be Very Complicated
One problem with functional neuroimaging is the interpretation of information. Very complex statistics are used to generate statistical maps that can translate to the likelihood of brain activity. The statistical maps can be very different based on the perimeters that the scientist uses. Even when a conclusive result is found, for example, a brain region definitely is active on EEG in a certain time frame, or an area of the brain lights up in fMRI, it remains unclear exactly what that means in terms of cognitive function.
Neuroimaging Data Remains Difficult to Interpret
Because of the limitations of human subjects research, non-invasive functional neuroimaging is one of our very few windows into how the human brain works. It’s definitely not something to take for granted, but it should also not be taken as a form of gospel regarding how our brains work. In reality, brain function is very complex, and a single study, or even a series of studies, conducted in a lab could never replicate the real-world activity of our brains. Therefore, especially given the limitations discussed above, all scientific studies involving brain function and brain imaging should be interpreted with a pinch of salt.
Noninvasive human brain imaging kicked off in the 1920s with the development of the EEG. Since then, scientists and clinicians alike have used human brain imaging to look at brain activity without using surgical means. While noninvasive brain imaging technologies are not without their limitations, they have also ushered in a new era of brain-related research and development.
Innovations like Elon Musk’s Neuralink seek to take brain activity measurement, such as that derived from the above methods, one step further. Neuralink seeks to not only directly record from individual brain cells in the human brain, a stated goal of the device is also to control brain activity by directly interfacing with the brain’s cells. These technologies rely heavily on things we have learned about the human brain from brain imaging and as such remain in their infancy.
It has been said that the brain is the final frontier of science. Thanks to functional neuroimaging, and growing interest in new brain implant technologies like Neuralink, as well as federal investments such as the NIH BRAIN Initiative, we will be exploring this frontier in much more detail in the coming years.