Electroencephalography (EEG) and magnetoencephalography (MEG) can track how different brain regions talk to each other with millisecond precision — a timing resolution no other noninvasive human brain-imaging method matches.
The research
In a book chapter posted July 20, 2026 to arXiv (q-bio.NC), University of Southern California researchers Richard Leahy and Takfarinas Medani walk through the full methodological toolkit for measuring brain networks with EEG and MEG. The chapter covers the physics behind both signals and why they complement each other: EEG picks up electrical currents, while MEG measures the tiny magnetic fields those currents produce, giving slightly different views of the same underlying activity.
The authors then tackle two classic problems. The forward problem asks how a known brain source would appear at the scalp sensors; the inverse problem asks the reverse — inferring where in the brain a measured signal came from. Getting this right requires subject-specific head models, accurate anatomy, and careful source reconstruction. Without that care, signals leak across sensors, a problem called volume conduction, which can create fake connectivity that looks real.
Leahy and Medani review the most common ways researchers measure connectivity between brain regions:
- Coherence and phase synchronization — do two regions rise and fall together in rhythm?
- Amplitude envelope correlation — do slow fluctuations in signal strength track each other?
- Granger causality and transfer entropy — does activity in one region predict activity in another?
- Dynamic causal modeling — a hypothesis-driven approach that tests how regions influence each other.
They also highlight end-to-end analysis pipelines, especially Brainstorm, an open-source software package widely used for reproducible EEG and MEG research. Emerging directions include time-varying connectivity (how networks shift moment to moment), cross-frequency interactions, and network-level analyses applied to health and disease.
Why it matters
Your brain's cognitive abilities — attention, memory, language — depend on rapid coordination between regions, not on any single area working alone. EEG and MEG let researchers observe that coordination as it happens, in real time. If you have ever wondered why some tasks feel "easier" for you, network-level measures are part of the answer: individual differences in how efficiently regions synchronize predict performance on attention and memory tasks. The methodological care described here matters too, because sloppy connectivity analysis can produce findings that look exciting but don't replicate — a persistent issue in cognitive neuroscience.
What you can do
You do not need a lab to benefit from this research. When you read brain-training or neurofeedback claims, ask which measure was used and whether volume conduction was addressed — that detail separates solid work from noise. Keep your brain's networks active with varied challenges: learn a new skill, switch between tasks deliberately, and get consistent sleep, which supports the oscillatory rhythms these methods record.
Source: arXiv q-bio.NC
Curious about your own brain? Take our free adaptive IQ test or try 306 brain training levels.