Brain Projects Plot Neural Connections

Abstract

Although BAM will study neuronal activity in the brain, the HCP is providing a static anatomy of the brain. In 2010, the NIH awarded $40 million to start the HCP. One project, dubbed the microconnectome, received $8.5 million over 3 years to examine the synapses and cells down to the resolution of 100 cubic microns. The larger project is led by co-principal investigators David Van Essen, Ph.D., a neurobiologist at Washington University in St. Louis, Missouri, and Kamil Ugurbil, Ph.D., a neuroscientist and MRI expert at the University of Minnesota. Van Essen and Ugurbil were awarded $30 million over 5 years to chart long connections that criss-cross white matter into gray matter in the cerebral cortex.Before any scans could happen, Ugurbil spent several years working on fMRI and diffusion tensor imaging scanners so that the resolution would be doubled over standard MRIs. Now, scans are underway at Washington University on healthy human adults. The plan is to scan 1200 healthy adults targeting 300 twin pairs, plus brothers and sisters of those pairs. About half of the twin pairs will be identical twins and the other half will be dizygotic or fraternal twins.Thus far, the twin studies show that the differences in brain structure from one human to another are striking, especially in cerebral cortical folds, even in the same person between the left and right hemispheres.Van Essen had spent so many years mapping these connections by hand while scrutinizing the visual systems in macaque monkeys that he lost a bet once the HCP twin data started to emerge. He was absolutely certain that he would be able to judge the maps so well by eye that he could tell apart the maps of study subjects just by looking at folding patterns.“I couldn’t do better than chance,” says Van Essen.The HCP has been compared to a view of Earth at 20,000 feet, and Van Essen says that is an accurate comparison. Think of looking at the earth’s surface and seeing geographic wrinkles caused by mountains, valleys, river channels, and oceans. Then, superimpose a map of the social and political subdivisions that also are routinely displayed on maps of the earth. Now, switch the analogy to human brains. “First, humans don’t have one brain, we have billions of brains, and the Human Connectome Project aspires to study 1200 of them to get as good a reading as we realistically can, but we are not looking at a smooth spherical surface, we are looking at highly convoluted surface,” says Van Essen.Aspiring to identify functionally significant subdivisions that are equivalent to political, social, and cultural subdivisions of Earth’s surface means that researchers need to do that mapping in a way that copes with the complexity of the brain’s convolutions. Making matters more complex is the fact that those convolutions themselves bring biases of their own to imaging methods used to study brain connectivity and it’s a continuing challenge to address those biases.“Diffusion imaging gives a distorted picture because that picture makes it look like neurons on the crowns of gyral folds are getting the bulk of action, and thus are the neurons doing the most talking or being talked to the most. This is an artifact,” Van Essen says. “However, understanding and eliminating are two different things.”In March 2013, Van Essen and Ugurbil released the first data set that encompassed brain imaging scans and behavioral information on 68 healthy adult volunteers. The team plans to release data quarterly until the 1200 volunteers are scanned.Data emerging from HCP will “almost certainly” converge with BAM on a scientific level, says Greenspan. “They have to. Data from each will be of enormous value to the other in both directions,” he says.It’s unclear as to how quickly data will emerge from BAM as the first request-for-proposals from the NIH is expected in the fall of 2013. “There is some criticism we don’t know what we want to measure,” says Peterka. “My answer is, if we don’t build this and start measuring, it will never get done.”