Brain Imaging Techniques: How Scientists See Inside the Brain

Structural Imaging: CT and MRI

Computed tomography (CT) scanning uses X-rays taken from multiple angles to construct cross-sectional images of the brain. CT scans can be acquired in seconds, making them invaluable in emergency settings for detecting hemorrhages, skull fractures, and large tumors. However, CT provides relatively low contrast between different soft tissue types and involves exposure to ionizing radiation, which limits its use for repeated imaging and research applications.

Magnetic resonance imaging (MRI) produces far more detailed structural images by exploiting the magnetic properties of hydrogen atoms in brain tissue. The scanner generates a powerful magnetic field that aligns hydrogen nuclei, then applies radiofrequency pulses that temporarily disrupt this alignment. As the nuclei return to their equilibrium state, they emit signals that vary depending on the tissue type, allowing the scanner to distinguish between gray matter, white matter, and cerebrospinal fluid with millimeter-level spatial resolution. Different MRI sequences can highlight different tissue properties: T1-weighted images provide excellent anatomical detail, T2-weighted images are sensitive to fluid and pathology, and diffusion tensor imaging (DTI) maps the orientation of white matter fiber tracts by tracking the directional movement of water molecules along myelinated axons.

Functional MRI

Functional MRI (fMRI) measures brain activity indirectly by detecting changes in blood oxygenation associated with neural activity. When neurons in a brain region become active, local blood flow increases within seconds to deliver additional oxygen and glucose. The blood-oxygen-level-dependent (BOLD) signal exploits the fact that oxygenated and deoxygenated hemoglobin have different magnetic properties, producing detectable changes in the MRI signal. This neurovascular coupling allows researchers to identify which brain regions are active during specific cognitive tasks, sensory stimulation, or emotional experiences.

fMRI provides good spatial resolution (typically 2 to 3 millimeters) but limited temporal resolution (typically 1 to 2 seconds), because the hemodynamic response to neural activity peaks several seconds after the neurons fire. Despite this limitation, fMRI has become the dominant tool in cognitive neuroscience, generating thousands of studies that have mapped the neural correlates of perception, attention, memory, language, emotion, and decision-making. Resting-state fMRI, which measures spontaneous fluctuations in brain activity while participants rest quietly in the scanner, has revealed the organization of the brain into functionally connected networks that persist even in the absence of external tasks.

Electroencephalography

Electroencephalography (EEG) measures the electrical activity of the brain through electrodes placed on the scalp. The signals detected by EEG represent the summed postsynaptic potentials of millions of cortical neurons firing in synchrony. EEG provides excellent temporal resolution, capturing neural dynamics at the millisecond timescale, but has limited spatial resolution because electrical signals become blurred as they pass through the skull and scalp. EEG is widely used in clinical settings to diagnose epilepsy, monitor sleep stages, assess depth of anesthesia, and detect brain death.

Event-related potentials (ERPs) are EEG signals time-locked to specific sensory, cognitive, or motor events and averaged across many repetitions to extract consistent neural responses from background noise. Well-characterized ERP components include the P300, a positive voltage deflection occurring roughly 300 milliseconds after an unexpected or significant stimulus, which reflects attentional processing and working memory updating, and the N400, a negative deflection associated with semantic processing of language. EEG and ERP methods are inexpensive, portable, and noninvasive, making them practical for research with populations including infants and patients who cannot tolerate the confined space of an MRI scanner.

Positron Emission Tomography

Positron emission tomography (PET) uses radioactive tracers injected into the bloodstream to measure metabolic activity, neurotransmitter binding, or protein deposition in the brain. The most common tracer, fluorodeoxyglucose (FDG), is taken up by metabolically active neurons, allowing PET to map brain regions with high or low glucose consumption. Specialized tracers can bind to specific neurotransmitter receptors, transporters, or abnormal proteins, providing information about neurochemical function that no other imaging method can offer.

PET has proven particularly valuable in clinical neurology for detecting the amyloid plaques and tau tangles of Alzheimer disease years before symptoms appear, for localizing seizure foci in epilepsy patients being evaluated for surgery, and for assessing dopamine system function in movement disorders. The main limitations of PET are its requirement for radioactive tracers (which limits repeated scanning), relatively low spatial resolution compared to MRI (typically 4 to 6 millimeters), and high cost of producing short-lived radioisotopes, which often requires an on-site cyclotron.

Magnetoencephalography

Magnetoencephalography (MEG) detects the extremely weak magnetic fields produced by electrical currents in active neurons, using highly sensitive superconducting quantum interference devices (SQUIDs) housed in a magnetically shielded room. Like EEG, MEG provides millisecond temporal resolution, but it offers better spatial localization because magnetic fields are less distorted by the skull and scalp than electrical signals. MEG is particularly useful for mapping the timing and location of sensory and motor processing, presurgical mapping of eloquent cortex in epilepsy and tumor patients, and studying the rapid dynamics of neural oscillations.

A key advantage of MEG is its sensitivity to the orientation of neural current sources, which complements the information provided by EEG. While EEG is most sensitive to radially oriented sources, MEG preferentially detects tangentially oriented sources in cortical sulci. Combining MEG with structural MRI allows researchers to localize neural activity with both high temporal precision and reasonable spatial accuracy, providing a more complete picture of brain dynamics than either method alone.

Emerging and Advanced Techniques

New imaging technologies continue to expand the capabilities of brain research. High-field MRI at 7 Tesla and above provides submillimeter resolution that can resolve individual cortical layers and small subcortical nuclei. Functional near-infrared spectroscopy (fNIRS) measures cortical hemodynamic changes using infrared light transmitted through the skull, offering a portable and relatively inexpensive alternative to fMRI that can be used during natural movement and social interaction. Optogenetics, while not a traditional imaging method, uses light-sensitive proteins to activate or silence specific neural populations in animal models, allowing researchers to establish causal relationships between neural activity and behavior that correlational imaging methods cannot provide.

Connectomics, the comprehensive mapping of neural connections, has emerged as a major frontier in brain imaging. Diffusion MRI tractography maps white matter pathways in the living brain, while serial-section electron microscopy reconstructs complete wiring diagrams at synaptic resolution in small brain volumes. The Human Connectome Project and related initiatives aim to map the full set of neural connections in the human brain, providing an anatomical foundation for understanding how brain structure gives rise to function.

Choosing and Combining Imaging Methods

Each brain imaging technique offers a different trade-off between spatial resolution, temporal resolution, invasiveness, and the type of information provided. Researchers frequently combine multiple methods to overcome the limitations of any single technique. Simultaneous EEG-fMRI recording, for example, captures both the high temporal resolution of electrical activity and the precise spatial localization of hemodynamic responses, providing a more complete picture of brain dynamics than either method alone. PET combined with MRI in integrated PET-MRI scanners allows simultaneous acquisition of metabolic and structural information in a single scanning session.

The choice of imaging method depends on the specific research or clinical question. Diagnosing structural brain abnormalities typically requires MRI for its superior soft tissue contrast. Investigating the rapid timing of cognitive processes calls for EEG or MEG. Mapping the spatial distribution of brain activity during complex tasks is best accomplished with fMRI. Assessing neurochemical systems requires PET with specific radioligands. Understanding which method provides the most relevant information for a given question is a fundamental skill in modern neuroscience, and the continued development of multimodal imaging approaches promises to provide increasingly comprehensive views of brain structure, chemistry, and function.