A microstate is the exact, detailed configuration of every individual part in a system at a single moment in time. The term shows up in two very different fields: statistical mechanics (physics and chemistry) and neuroscience. In physics, it describes the precise positions and speeds of every particle in a gas or solid. In neuroscience, it describes a brief, stable pattern of electrical activity across the brain’s surface. Both uses share the same core idea: a snapshot of a complex system captured at its most granular level.
Microstates in Physics and Chemistry
Imagine a box of gas. You can describe that gas in broad terms: its temperature, pressure, and volume. Those broad descriptions are called macrostates. A microstate, by contrast, specifies the exact position and momentum of every single particle in the box at one instant. For a system of N particles, that means tracking N positions and N momenta simultaneously.
Here’s the key insight: many different microstates can produce the same macrostate. A room might be 72°F, but the individual molecules bouncing around inside it could be arranged in a staggering number of ways that all result in that same temperature reading. Statistical mechanics is the branch of physics that bridges these two levels of description, connecting the behavior of trillions of individual particles to the bulk properties we can actually measure with a thermometer or pressure gauge.
Counting Microstates
The number of microstates corresponding to a given macrostate is often represented by the symbol W (or sometimes Ω). For simple systems, you can calculate W directly. Take a row of 36 magnetic spins that can each point up or down. If the macrostate is “13 spins pointing up,” the number of microstates is the number of distinct ways to choose 13 out of 36 positions, which works out to 36! divided by (13! × 23!). That’s already millions of arrangements for a tiny system.
For a realistic gas with something like 10²³ particles, the number of microstates is astronomically large. This is where entropy comes in. The Boltzmann entropy formula ties the two concepts together: S = k_B × ln(W), where S is entropy, W is the number of microstates, and k_B is Boltzmann’s constant (about 1.38 × 10⁻²³ joules per kelvin). In plain terms, entropy is a measure of how many microscopic arrangements are compatible with what you observe at the large scale. More microstates means higher entropy, which is why systems naturally drift toward their most probable macrostate: it simply has the most microstates behind it.
Microstates in Neuroscience
In a completely separate field, “microstates” refers to brief, recurring patterns of electrical activity measured across the scalp using EEG (electroencephalography). During an EEG recording, the topographical map of voltage across your head doesn’t shift smoothly and continuously. Instead, it holds a stable pattern for roughly 80 to 120 milliseconds, then snaps to a different pattern. Each of these quasi-stable periods is called a microstate.
Think of it like a slideshow rather than a movie. Your brain’s electrical landscape clicks through a rapid sequence of distinct frames, each one lasting about a tenth of a second before transitioning to the next. Researchers believe these microstates represent the basic building blocks of conscious thought, the smallest units of cognitive processing that can be captured from the scalp.
The Four Canonical Classes
Decades of research have identified four standard microstate patterns, labeled A through D. Each class has a characteristic spatial layout on the scalp and has been linked to a different mode of mental activity:
- Class A: associated with verbal and phonological processing
- Class B: associated with visual processing
- Class C: associated with subjective interoceptive and autonomic processing (your brain monitoring internal body states)
- Class D: associated with attention reorientation (shifting focus to something new)
These four classes account for a large share of the variance in resting-state EEG data, meaning that at any given moment, your brain’s electrical topography closely resembles one of these four templates. More recent work has expanded the model to seven classes (adding E, F, and G), with classes like F and G linked to internal processing and body-related sensory functions. However, the four-class model remains the most widely used.
How Microstate Sequences Reveal Brain Function
It’s not just which microstates appear that matters. The order in which they occur, how long each one lasts, and how often each class shows up all carry information about cognitive function. Researchers analyze transition probabilities: the likelihood that microstate A will be followed by C rather than B, for example. These sequences are not random. They follow scale-free dynamics, meaning the same statistical patterns appear whether you zoom in on short time windows or long ones.
Sleep deprivation offers a clear illustration. After going without sleep, people show a higher probability of transitioning from microstates A and B into C, and a lower probability of transitioning from C and D into B. Researchers interpret this as preserved bottom-up sensory processing (the brain still responds to incoming signals) but weakened top-down control (your ability to deliberately direct attention degrades). The temporal sequence of microstates essentially provides a window into how spontaneous brain activity reorganizes under different conditions.
Microstates and Neurological Conditions
Some of the most striking findings involve psychiatric and neurological disorders. In schizophrenia, one of the most consistent observations across multiple studies and meta-analyses is that microstate class C, the pattern tied to internal body monitoring, shows increased duration and increased time coverage compared to healthy controls. Meanwhile, class B (visual processing) shows decreased occurrence. This pattern holds across patients with both high and low levels of positive symptoms (like hallucinations) and negative symptoms (like social withdrawal), suggesting it reflects something fundamental about how the condition alters brain dynamics rather than a side effect of any single symptom.
Alzheimer’s disease produces a different signature. Patients show prolonged duration of microstates C and D, along with reduced occurrence of microstate B and a decrease in overall microstate frequency. After rigorous statistical correction, the prolonged duration of microstate C remained the most robust finding. This slowing of microstate dynamics may reflect the broader cognitive deceleration characteristic of the disease, and researchers are exploring whether microstate patterns could help predict the degree of amyloid protein buildup in the brain.
How EEG Microstates Are Identified
Extracting microstates from raw EEG data requires clustering algorithms that group similar topographical maps together. The traditional approach uses modified K-means clustering: the algorithm identifies the handful of spatial patterns that best explain the full recording, then assigns each time point to the pattern it most closely resembles. Newer methods incorporate more sophisticated techniques, including Riemannian geometry (a way of measuring distances between complex spatial patterns) and deep autoencoders, a type of neural network that compresses high-dimensional data into a simpler representation before clustering. These advanced approaches are particularly useful for task-related EEG, where brain activity is more variable than during rest.
Regardless of the algorithm, the output is the same: a continuous EEG recording gets reduced to a color-coded sequence of microstate labels, like a string of letters (A, B, C, D, A, C, D, B…) that researchers can then analyze for timing, frequency, and transition patterns.
Connecting the Two Definitions
The physics and neuroscience uses of “microstate” are independent, but they share a deep conceptual parallel. In both cases, complex systems are described at two levels. The micro level captures fine-grained detail (particle positions, millisecond-by-millisecond brain topography), while the macro level captures overall properties (temperature, cognitive state). And in both cases, the number and behavior of microstates tell you something fundamental about the system: how disordered a gas is, or how a brain organizes its moment-to-moment activity. The word means the same thing at its root. It’s the smallest complete description of a system’s state at a single point in time.