The Human Connectome
Effort to relate brain connectivity to individual capabilities will establish a new baseline for future studies

By Michael C. Purdy

Learn More

The Human Connectome Project


What makes us think and feel and remember? Why do people behave differently or have varied learning styles? What makes people unique? The answers can be found in the neural networks firing within our brains.

The Human Connectome Project (HCP) is a $30-million, five-year effort to comprehensively map human brain circuitry — and its variability — in 1,200 healthy adults using advanced, noninvasive neuroimaging methods. Washington University, the University of Minnesota and Oxford University are leading a 10-institution consortium involving more than 100 investigators and staff.

The project is gathering and analyzing data, and making its results freely available to the scientific community, in the hope of sparking future discoveries — shedding light on healthy brain function and, ultimately, informing doctors’ ability to diagnose and treat disorders.

HCP by the numbers
$30 million
5 years
1,200 healthy adults
300 twin pairs &
non-twin siblings
10 institutions
100 investigators
Projected Dataset
1 petabyte
Equivalent to
1 million gigabytes, or
58,000 DVD movies
Open access Freely available

It took centuries to arrive at today’s detailed views of Earth as speculative imaginings yielded to exploration and technology. While cartographers only had one Earth to map, HCP collaborators are creating maps that may be useful in understanding billions of human brains. Researchers are exploring the similarities and differences that make us who we are, giving us all our unique mix of personality quirks, mental traits and individual identities.

In short, the HCP is conducting the most complex and ambitious mapping ever attempted: charting the intricacies of the living, working human brain. Co-principal investigator David Van Essen, PhD, appreciates the challenges as well as the limits of this historic endeavor.

“In comparison with our historical efforts to chart the Earth’s surface, the HCP will, in effect, take us from an 18th-century representation of the cortical landscape to a 19th-century representation,” Van Essen said.

Why this study is so challenging

Researchers seek to interpret data with respect to the human brain’s organizational features and structural dimensions. However, the size, function and connectivity of the parts vary so greatly among individuals that making comparisons and establishing averages becomes a key challenge. Each critical area may connect with dozens of other areas and the connection strengths differ over many orders of magnitude.

The outer cerebral cortex (gray), occupies about 80% of brain volume but contains only about 20% of the brain’s 85 billion neurons. Its convoluted surface has 150–200 critical areas of study, which vary in size twofold or more across the normal adult population.

Some 80% of the neurons reside in the cerebellar cortex (gold). This region of the brain also varies highly among individuals.

Subcortical structures (blue) occupy the remaining approximately 10% of brain volume but contain only about 1% of the neurons.

Understanding complexity

Scientists admit the human brain remains a mystery. Each human brain, weighing about 3 pounds, contains about 85 billion neurons, or nerve cells — more than 12 times the number of people on Earth — which transmit information across roughly 150 trillion cell-to-cell connections known as synapses. In nearly all brain disorders, there’s something wrong with these connections, though, in most cases, scientists don’t yet know exactly what has gone awry.

In the last two decades, an explosion of new brain imaging methods has yielded vast amounts of information.

Functional magnetic resonance imaging (fMRI), developed in the 1990s, opened the door to direct observation of cognitive activities. By detecting changes in blood oxygenation and flow, fMRI produces activation maps showing which parts of the brain are involved in specific mental processes. Using fMRI, researchers are scanning participants while they perform tasks and at rest.

Diffusion MRI gives insight into how regions of the brain are physically wired together into networks that communicate electrical signals through the white matter.

Structural MRI scans clearly show the physical features of the participants’ brains; these features vary greatly among individuals.

See “Making connections,” below, to learn more about how HCP data is being gathered from the human test subjects.

Holding bookends that model the human brain: Washington University HCP lead investigators Deanna Barch, PhD, the Gregory B. Couch Professor of Psychiatry and chair of the Department of Psychology, and David Van Essen, PhD, the Alumni Endowed Professor of Neurobiology. The HCP’s co-principal investigators are Van Essen and Kamil Ugurbil, PhD, a professor at the University of Minnesota.
Photo by Robert Boston

Wrinkles unraveled

Much of the brain’s uniqueness is found in the cerebral cortex, the wrinkled outer layer of the brain often known as the gray matter.

“The cortex is like a sheet that is crumpled and folded up to fit inside the skull,” Van Essen explained. “This folding pattern varies from one person to another, and that makes it really hard to just look at a brain and say, ‘okay, how is it organized?’”

Scientists already knew that the wrinkles, or folding patterns, of the cerebral cortex are what they call “partially heritable.” A glance at the brains of a pair of twins reveals major differences, but closer study uncovers structural elements that strongly suggest common genetic influences.

Prior to the Connectome, Van Essen was a leader in the application of visualization techniques that effectively inflate the crumpled sheet into a spherical beach ball, making it easier to match up significant features from one brain to the next.

Through the advanced scanning and data-analysis techniques developed and applied by HCP, scientists have created another tool for standardizing these comparisons: maps of myelin, the protective white material that wraps many nerve cell branches. Connectome scientists devised ways to use aspects of this material like the mountains, rivers and other landmarks that helped earthly mapmakers get their bearings.

“We think there are about 150 to 200 cortical areas — analogous to the political subdivisions of Earth’s surface — that are critically important for understanding brain function and connectivity,” Van Essen said. “Myelin maps give us markers for an interesting subset of this mosaic of areas. We can see these markers in individual subjects and watch how their location jumps around from one subject to the next.”

One of Van Essen’s MD/PhD students, Matthew Glasser, is leading an effort to create a new master map of the cortex based on combining myelin maps with other HCP data.

They expect to unveil the map this year.


At rest Average functional connectivity is collected on subjects while at rest in the MRI scanner. Yellow regions are functionally connected to a “seed” location in the right hemisphere, whereas regions in blue and purple are weakly connected or not connected at all.

Myelin Red and yellow regions have high myelin content compared to other colors. Mapping the myelin on the cortex indicates functionally distinct brain regions — an important step toward understanding how brain networks form dynamically to do tasks.

Story vs. math Participants are asked to listen attentively to either a story or a calculation (e.g. “16 minus 5 plus 18 equals”) and answer a question about what they heard. Stories activate yellow and red regions, whereas math affects green and blue regions.


Future implications

HCP scientists made their fourth large data release earlier this year, doubling the data they have made available to the public.

The release included scanning, demographic and behavioral data from more than 500 subjects. The information is organized in a web-accessible database called ConnectomeDB. A software application, Connectome Workbench, allows users to view and compare many types of imaging data, both produced by HCP and by other labs.

“These maps are going to have a major impact on our understanding of the healthy adult human brain, and that impact is already starting to be felt,” Van Essen, said. “Not only have we published many papers on what we’re learning about the brain from the data-gathering process, there also already have been a handful of independent investigators who have published papers based on HCP data. We weren’t expecting that this early in the process.

“We’re approaching the point where we have enough data to really sink our teeth into these questions and are starting to do some serious science on what factors influence the folding of the cortex,” he said.

Lead investigator Deanna Barch, PhD, said that to comprehend how structural or functional connectivity goes awry, researchers must first establish the parameters of normal connectivity.

“You need a baseline to work from,” she said. “Down the road, the HCP will become a benchmark enabling us to see how abnormalities contribute to or result from disease. It will suggest better ideas about disease prevention or the means of early intervention.”

The ‘state’ of the art in brain mapping

Imagine mapping population centers by tracking telecommunications activity — say, grouping at least 100,000 phone calls from one location to another. This would reveal Kansas City and St. Louis as two hot spots in Missouri, but would not identify individual callers. HCP mapping, somewhat similarly, monitors the overall activity of more than 100,000 neurons at a time. This allows researchers to better understand the brain’s functional hot spots and their connectivity, providing the most detailed and accurate whole-brain portrait yet developed.

Tucker Hartley and Owen Footer, HCP research assistants, perform some behavioral tasks used by the HCP to measure each subject’s abilities, such as gait speed (top), odor identification (middle), and taste (bottom). At right, Footer prepares Hartley for a functional MRI scan in which the subject indicates choices with a button box.
Photo by Robert Boston

Making connections

HCP researchers are studying twins and their non-twin siblings, a total of 1,200 subjects, to better understand how genes affect the wiring of the brain: What aspects of circuitry are inherited and what aspects are shaped by experience?

To reflect the true normal range of brain connections, study volunteers come from all racial, ethnic, economic and educational backgrounds.

Study volunteers spend about four hours over two days in a customized MRI, sometimes lying quietly and sometimes performing tasks. All the while, the machine is recording patterns of brain activity. A participant might move her right hand, for instance, and a section of the left motor cortex lights up. The testing is replicated in hundreds of subjects. For something more complicated, such as a working memory test, several brain regions in the front and back light up together, working in a coordinated way.

“We are trying to understand which parts of brains are active when people are processing information, doing math or remembering things,” explained Barch, who is coordinating data collection at Washington University.

In addition to the brain scans, research volunteers undergo a battery of cognitive, physical fitness and psychological tests. They spend about six hours participating in wide-ranging activities — from tasting mysterious liquids on a cotton swab to running around traffic cones in a corridor. These tests, which look at intelligence, sensory abilities and emotional state, among other aspects, bring added dimension to the imaging data. For instance, do differences in brain circuits have any bearing on high or low IQ?