3.3 Neuroimaging Techniques

Neuroimaging Principles and Applications

Neuroimaging techniques let researchers observe brain activity while people perform cognitive tasks. Tools like fMRI, EEG, and PET each measure different aspects of brain function, and understanding what each one actually detects (and what it misses) is central to evaluating research in cognitive psychology.

Principles of Neuroimaging Techniques

Each neuroimaging method works through a fundamentally different mechanism, which is why they produce different kinds of data.

Functional Magnetic Resonance Imaging (fMRI)

fMRI tracks changes in blood oxygenation across the brain. When a brain region becomes active, its neurons consume more oxygen, triggering increased blood flow to that area. This is called the BOLD signal (Blood-Oxygen-Level-Dependent). Strong magnetic fields and radio waves detect these oxygenation changes and produce detailed 3D images of the brain.

Common applications include:

• Mapping which brain regions are involved in specific cognitive tasks (e.g., working memory, decision-making)
• Studying how brain regions communicate within networks (e.g., the default mode network)
• Investigating cognitive disorders and brain plasticity, such as reorganization after stroke

Electroencephalography (EEG)

EEG records the brain's electrical activity through electrodes placed on the scalp. These electrodes pick up voltage fluctuations produced by ionic currents flowing within large populations of neurons. Because electrical signals travel almost instantly, EEG captures changes on a millisecond timescale.

Common applications include:

• Studying brain wave patterns tied to different cognitive states (e.g., alpha waves during relaxation, theta waves during focused attention)
• Investigating sleep stages and sleep disorders through characteristic wave patterns (e.g., REM sleep)
• Monitoring real-time brain responses during cognitive tasks like attention switching or language processing

Positron Emission Tomography (PET)

PET works by injecting a small amount of radioactive tracer into the bloodstream. As the tracer is taken up by active brain tissue and decays, it emits positrons that collide with electrons, producing gamma rays. Detectors surrounding the head pick up these gamma rays and map where metabolic activity is occurring.

Common applications include:

• Studying neurotransmitter systems (e.g., dopamine pathways in Parkinson's disease)
• Measuring brain glucose metabolism (e.g., reduced metabolism in Alzheimer's disease)
• Locating abnormalities such as epilepsy foci or brain tumors

Strengths vs. Limitations of Neuroimaging

This comparison matters because the method a researcher chooses shapes what conclusions they can draw.

A key tradeoff to remember: fMRI tells you where with precision, EEG tells you when with precision, and PET tells you what chemicals are involved. No single technique does everything well, which is why researchers sometimes combine methods (e.g., simultaneous EEG-fMRI).

Interpretation of Brain Activation Patterns

Reading neuroimaging results requires understanding what those colorful brain maps actually represent.

Activation maps show regions where neural activity increased during a task compared to a baseline or control condition. The colors typically indicate intensity of activation, not a simple on/off switch. A region shown in red doesn't mean it's the only area involved; it means activity there was statistically greater than during the comparison condition.

Functional specialization refers to the idea that certain brain regions are particularly associated with specific cognitive processes. Broca's area, for example, is strongly linked to language production, and the hippocampus to memory formation. But most cognitive tasks recruit distributed networks, meaning multiple regions work together. Working memory, for instance, involves prefrontal cortex, parietal regions, and other areas acting in coordination.

Individual differences also shape activation patterns. Age, expertise, and cognitive ability can all influence which regions activate and how strongly. Two people performing the same task may show noticeably different patterns, which is why group-level analyses and adequate sample sizes matter.

Evaluation of Neuroimaging Studies

When you're reading or critiquing a neuroimaging study, these are the main questions to ask:

1. Experimental design: Were the control conditions appropriate? A poorly chosen baseline can make results misleading. The task should clearly isolate the cognitive process being studied.

2. Statistical analysis: Neuroimaging data involves thousands of data points (one per brain voxel), so researchers must correct for multiple comparisons. Without this correction, you'll get false positives, meaning brain regions that appear active by chance. Common analysis methods include the general linear model (GLM) and multivariate pattern analysis (MVPA).

3. Sample size and statistical power: Many early neuroimaging studies used very small samples (sometimes fewer than 15 participants), which increases the risk of both false positives and false negatives. Larger samples produce more reliable and replicable findings.

4. Reverse inference: This is a common reasoning error in neuroimaging. Just because a brain region activates during a task doesn't mean the mental process you associate with that region is occurring. For example, the amygdala activates during fear, but it also activates during other emotional states and even during some non-emotional tasks. Inferring "fear" just because the amygdala lit up is reverse inference, and it's logically flawed.

5. Generalizability: Lab tasks are often simplified versions of real-world cognition. Consider whether findings from a controlled experiment (e.g., memorizing word lists in a scanner) extend to how memory works in everyday life.

6. Replication: Have the findings been reproduced by other labs, with different participants, or using different paradigms? Robust findings hold up across multiple studies. Results that appear only once, in one sample, should be interpreted cautiously.