With New BRAIN Initiative Grant, Jon Polimeni Seeks More Faithful Measure Of Neuronal Activity

Gary Boas
December 8, 2016


Investigator Jon Polimeni is working to improve the neuronal specificity of fMRI. Photo by of Matti Hamalainen.

Functional MRI has provided unprecedented insights into the brain mechanisms underlying a range of behaviors. It has done this by tracking local changes in blood flow and oxygenation in the brain that accompany neuronal activity. But these changes are only a proxy for increases in neuronal activity, and it’s not always clear how precise of a proxy they are. For this reason, the very phenomenon that makes fMRI possible is also often viewed as a fundamental limitation.

Researcher Jon Polimeni is looking to change this. Polimeni, Director of Ultra-High Field Imaging at the MGH Martinos Center for Biomedical Imaging and an Assistant Professor of Radiology at Harvard Medical School, was recently awarded a grant through the BRAIN Initiative at the National Institutes of Health to address the potential mismatch. He proposes to develop, through a combination of basic science research and technology development, a means to more closely estimate changes in neuronal activity with fMRI.

In a recent conversation he explained what will make this possible. “Recent studies have demonstrated that some components of the blood flow response to neuronal activity can be far more precise than what was commonly believed,” he said, “but the blood flow responses are only more faithful to neuronal activity at spatial scales smaller than currently available MRI voxel sizes.”

This suggests that fMRI can in fact provide a much more truthful representation of neuronal activity than is possible today—“if,” Polimeni said, “we can shrink our voxel sizes and extract these more specific components.”

Here’s where things get a little tricky. Voxel sizes—a voxel is a three-dimensional unit of measurement used in imaging, a volumetric pixel—are already pushing the limits of what modern technology allows. So making them smaller is, as they say, easier said than done. Motivated by the recent discoveries of the tight coupling between blood flow changes and neuronal activity described above, Polimeni and colleagues have recently developed new imaging technology to shrink the voxel sizes used in functional MRI. With this new technology for high-resolution imaging, they will map out the functional architecture of the working human visual cortex at unprecedented spatial scales.

The idea is to improve the neuronal specificity of functional MRI by linking detailed micro- and macroscopic maps of the human microvascular anatomy. In the proposed project, this will involve a three-step process: (1) performing microscopic, CLARITY-based imaging of human brain specimens and mapping the full microvasculature of the human visual cortex (CLARITY is a relatively new technique that makes brain tissue transparent through a series of chemical treatments); (2) making predictions about the patterns of changes in functional MRI signal based on these microanatomical maps; and (3) comparing these predictions to high-resolution measurements conducted with the Ultra-High Field 7 Tesla MRI scanner at the Martinos Center.

It is “an admittedly ambitious, multi-faceted project,” but the payoff could be tremendous, especially over time.

“Inferring neuronal activity from the functional MRI signal is a ‘holy grail’ of human neuroimaging, and it is an extremely challenging problem that we will likely not fully solve in this one project,” Polimeni said. “This study is designed to further our understanding of the relationship between neuronal activity and functional MRI, and push functional MRI to map out human brain function at unprecedented scales, to take steps towards this ultimate goal.”