What is lost when we study an 'average' brain? A map built from many people's scans shows common patterns. In one person, however, a network may consistently appear elsewhere. That difference raises the question of how to study shared features while preserving each brain's organization.
The problem with an average fingerprint
Imagine averaging thousands of fingerprints into one picture. Some common textures might remain, but a loop that appears in different places on different fingers could disappear. Caterina Gratton and Rodrigo M. Braga use this image to explain a problem in brain research. A map made by averaging many people's scans may show broad patterns while hiding the detailed arrangement of the networks in any one person's brain.
Functional magnetic resonance imaging, or fMRI, tracks changes in a blood-oxygen signal associated with brain activity. Researchers use the way signals vary together to infer networks of regions that tend to work together. Traditional studies often collect a limited scan from each participant and then align and average the results. The average is useful for finding features shared in roughly the same place. It is less useful when a meaningful feature is shifted, enlarged or divided differently from one person to another.
The review seeks shared principles while accounting for differences in each brain's organization. Its question is: which principles recur across people once we stop requiring every feature to occupy exactly the same coordinate? To answer it, researchers need a dependable map within each person first. Otherwise a blurred scan and a real individual difference may look the same.
Many hours with one person change what can be seen
The method called precision fMRI collects much more data from each participant, often across several sessions. One influential project scanned the same researcher more than a hundred times over a year; others followed small groups repeatedly. The purpose is to make each person's network map stable enough to inspect before comparing people. Repetition reduces the chance that a peculiar patch is only a momentary fluctuation or a short scan's noise.
With extensive data, the authors describe network variants: places where one individual's pattern of connections differs substantially from the group map. A region assigned to a control network in the average, for instance, may connect like part of the default network in a particular person. Variants can differ in location, shape and size. That is a spatial difference in measured network organization, not a finding that this person is deficient or possesses a special talent.
Researchers also ask whether the variant persists. In the studies reviewed, much of a person's network layout remained similar across days, weeks and different tasks; changes due to the immediate state were generally smaller. Some individual maps also matched the brain's response during a task better than the group-average map did. Agreement across repeated scans and different measures strengthens the case that these patterns are real, while leaving open what caused them and whether they affect daily life.
Finding the same network without identical coordinates
If a network moves from one person to another, how can researchers know that it is the same network? The answer cannot be merely that the spots overlap. Gratton and Braga describe several anchors: a broad cortical zone, a characteristic pattern of connections, the network's position relative to neighboring networks, its approximate size and its response during particular tasks. These clues work together much like recognizing a familiar street in different city plans.
The review illustrates this with two neighboring parts of what group maps had treated as the default mode network. In carefully mapped individuals, the authors distinguish networks called DN-A and DN-B. Their pieces can lie close together yet maintain different connections across the brain. In one person's map, a piece may not overlap the corresponding piece in another person's map at all. It can still be identified by the broader zone it occupies and by its relationships to other pieces.
This approach changes what ‘common’ means. Common organization need not consist of identical dots on a template. It can consist of a repeated arrangement: two networks appearing in characteristic zones and maintaining similar relations even as their exact borders vary. That is a stronger and more interesting claim than either ‘all brains are alike’ or ‘no two brains can be compared.’ It also requires disciplined criteria; without them, any unusual patch could be made to fit a favorite story.
Details that change a familiar brain picture
A sharper individual map can reveal organization blurred by a group average. In the lateral prefrontal cortex, precision studies find small, interleaved areas with different preferences, including regions linked more closely to visual or auditory attention. That complicates a simple picture in which the entire area performs one general-purpose job. It does not mean researchers have found a separate module for every thought; it means a broad label can conceal finer structure.
The motor cortex provides another example. The classic classroom drawing represents body parts progressing neatly along a strip of cortex. Individual connectivity maps reveal regions between more specialized movements that connect with broader networks for planning and action. Researchers call this the somato-cognitive action network. The finding adds a layer to the motor map: moving the body is not only a matter of placing isolated body parts on a strip.
The review also examines links between cortical networks and deeper brain structures, including the thalamus and basal ganglia. More detailed maps make it possible to ask whether a network in the cortex has a corresponding pattern below it. These examples illustrate the value of precision mapping for anatomy and theory. Their details are still being investigated and should not be translated directly into claims about a person's motivation, character or health from one scan.
Stable layout does not mean an unchanging brain
Much of a person's network layout looks stable across ordinary scanning sessions, but stability is not permanence. The authors review people whose brains developed after early injuries, sometimes with major changes in where motor or language networks were placed and less change in ability than the map might suggest. One case involved motor functions supported in locations shifted after a probable early stroke. Such cases reveal developmental flexibility; they are not a prediction that every early injury will be harmless.
Experience also leaves traces on different timescales. Studies of limb immobilization, drugs and hormonal changes suggest that the strength of communication among regions can shift over days. Repeated scans across menstrual cycles and pregnancy show how within-person measurements can detect change that a one-time group comparison may miss. The nature, size and persistence of these changes vary, so a fluctuation in connectivity should not automatically be called a rewiring of a whole network.
The authors distinguish a network's topography—where its parts lie—from the strength of the measured correlations among those parts. Layout often changes slowly and remains recognizable, while connection strength may vary more quickly with state or experience. Both kinds of variation can matter. The distinction also prevents a common error: a single scan taken on a difficult day cannot establish that someone's enduring brain organization has changed.
A personal map is not a personality test
Do these individual maps explain differences in thinking and behavior? Some links are promising. A person's resting network map can correspond to where the brain responds during a task. Researchers have also used aspects of network layout to predict some measured cognitive characteristics across samples. That makes the differences worth studying. But the prediction of behavior from brain scans is usually modest, and a correlation in a study is not a reading of an individual mind.
The review highlights a productive tension. Some people with strikingly unusual layouts function well; some exceptional performers have network distributions that appear surprisingly ordinary. Large studies often find weak relations between a single brain measure and a behavioral trait. Measurement error, the complexity of behavior and the brain's capacity to reach similar outcomes through different arrangements may all contribute. Precision in the image does not automatically produce precision in a psychological forecast.
Individual maps allow researchers to investigate features that an average image may obscure. Which spatial features reliably recur within a person? Which change after experience? Which predict a well-defined task result in an independent group? Which predictions survive when the sample grows? A personal map can reveal hidden biological organization without becoming a label for the person. Studying each brain in detail can help identify shared organization. That conclusion requires results that recur: a clear image does not by itself establish a claim about the brain.
The tradeoff is also practical. Scanning a small number of people for many hours makes fine details visible but does not tell us how frequent they are in a population. A much larger group gives breadth but may devote too little time to mapping each member accurately. The authors' examples suggest a sequence: discover a feature in deeply studied individuals, develop a measure that can identify it with less data, and then test that measure in broader, independent samples. Neither depth nor breadth can do the other's job alone.
What to keep
- An average brain is not every brain.
- Repeated scans improve individual maps.
- Precision creates new conceptual questions.
The nuance
What this story does not let us claim
Deep measurement of a few people trades sample breadth for within-person detail and requires demanding quality control.
Go deeper: concepts and optional reading practice
Reversible cards
Test the concept
Topic assessment
Check whether you can explain the idea, not merely recognize it.
Every answer is grounded in the explanation above. Some questions introduce new cases to test whether you can apply and analyze the ideas, not merely recall them.
Source & scope
Dense Phenotyping of Human Brain Network Organization Using Precision fMRI. Caterina Gratton & Rodrigo M. Braga. Annual Review of Psychology, vol. 77 (2026).
This editorial guide is an independent explanation. Consult the original publication for the studies, references and full context.
Open the original review ↗