Brain Scanning Just Got Very Good—and Very Unsettling

null
Image: Anastasia Yendiki and Viviana Siless/MGH/Harvard University

Seven years ago, the U.S. National Institutes of Health (NIH) decided to map all the connections in the brain. In 2010, the Human Connectome Project (HCP) was born. It has provided funding to the tune of $40 million to two collaborating consortia whose aim was to acquire and share high-resolution data of structural and functional connections in the human brain. The researchers have sought to understand, on a scale never before attempted, the neural pathways that make us human, and how changes in those pathways make us sick.

At a symposium yesterday at the NIH campus in Bethesda, Maryland, top researchers from the HCP came together to provide an update on the project’s achievements and future directions. To date, the consortia have released brain-scanning data from hundreds of individuals and that data has been used in more than 140 scientific publications. Perhaps even more importantly, the effort has produced impressive new tech, including unprecedented magnetic resonance (MR) hardware. Among the gadgets are high-powered scanners and customized head coils. In addition, there are legions of software for analyzing, visualizing and sharing the petabytes of neuroimaging data being generated.

The HCP project has also moved brain scanning into the realm of the feature film Minority Report by showing that a person’s brain activity is as unique as a fingerprint and that it can be used to identify a person with 99 percent accuracy. HCP data has also enabled researchers to use a brain scan to predict how a person will perform on an intelligence test and during a memory or reading task. “This may be a bit scary,” admitted Roderic Pettigrew, director of the National Institute of Biomedical Imaging and Bioengineering, during introductory remarks at the symposium.

One of the consortia, with collaborators at Washington University in St. Louis and the University of Minnesota, has now completed scanning the brains of 1,200 healthy adults. This week, they will upload the last of that data to the Internet so that it is freely available to neurologists all around the world. So far, more than 5,600 investigators have entered the public database, said project leader David Van Essen of WashU. Once there, they are able to access and download 9.5 petabytes of imaging data. The data set includes scans of brain activity while individuals are at rest and while doing tasks. It also contains structural scans with information on the size and shape of the folds of the brain in the cortex, the trajectories of local and long-distance neuronal fibers traversing the brain’s white matter, and more.

In order to collect of that data, the HCP team pioneered the use of a “multiband” approach, which involves scanning three sections of the brain at the same time instead of in series. This method has allowed radiologists to acquire high-resolution images up to 8 times as fast as they previously could using traditional MRI machines. “That gives us higher quality data; it also gives us a ton more data,” said Van Essen at the symposium.

The same technique can now be used in conventional imaging at hospitals, added Bruce Rosen of Massachusetts General Hospital, project leader of the other consortium. “This will have as profound an impact on how we do clinical, routine MR studies as any advancement in the last decade of MR imaging.”

Rosen’s group, a collaboration between Massachusetts General and the University of California at Los Angeles, has built a custom MRI scanner they call the “Connectom.” The machine, located at Massachusetts General, uses strong magnetic fields to track the movement of water through white matter in the brain. This diffusion MRI enables the visualization of brain networks with 10 times the level of detail possible with conventional MRI scanners. And it does the job so much faster. Getting a high-resolution 3-D image of the brain used to take 24 hours. Now it takes well under an hour, said Rosen. To achieve that kind of speed and resolution, the machine draws 22 megawatts of power—about the same as a nuclear-powered submarine. “It’s a lot of current and a lot of water cooling in one compact space,” said Rosen.

Now that the project has produced the intended amount of baseline data from healthy, middle-age adults, it’s time to look at connectivity in different types of brains. The next stage of the research will assess connectivity across the human lifespan, from infants to the elderly, as well as in patients manifesting various forms of human disease. At Stanford University, for example, a team is using HCP technology to study the anatomy of victims of traumatic coma to see if they can predict if and how a patient will improve. At Massachusetts General, radiologists are using the Connectom scanner to look at the microstructure of brain lesions in patients with multiple sclerosis.

The HCP may help us understand how human brains work so impressively, and yet are susceptible to devastating disorders, said Van Essen. “This is one of the grand challenges for the century, if not the millennium, ahead.”

Advertisement

The Human OS

IEEE Spectrum’s biomedical blog, featuring the wearable sensors, big data analytics, and implanted devices that enable new ventures in personalized medicine.

 
Senior Associate Editor
Eliza Strickland
Contributor
Emily Waltz
 
Contributor
Megan Scudellari
 

Newsletter Sign Up

Sign up for The Human OS newsletter and get biweekly news about how technology is making healthcare smarter.

Advertisement