Research Press Release

Fingerprinting the brain

Nature Neuroscience

October 13, 2015

The unique pattern of connections within a person’s brain can be used-like a fingerprint-to reliably identify individuals from a large group of people, reports a study published online this week in Nature Neuroscience. The study shows that this fingerprint-like pattern of connections can also be used to predict a person’s ability to think logically and quickly solve problems in unfamiliar situations.

Neuroimaging research has traditionally examined brain activity across different groups of people to better understand brain function across the general population. Studies that contrast two groups often ignore individual differences in brain activity, but whether the unique characteristics of a person’s brain connection profile are stable enough to identify an individual from a large group remained an open question.

Using data from 126 participants of the Human Connectome Project, Emily Finn and colleagues demonstrate that a connection profile obtained from one of six imaging sessions-two while participants rested and one each as they performed working memory, emotion, motor, and language tasks-can be used to identify an individual from a set of profiles obtained in a subsequent session. The authors found that they could successfully identify individuals across the resting and task-related sessions, indicating that the overall connectivity profile is intrinsic to the individual. They also found that connectivity profiles could predict fluid intelligence, a measure of the ability for quick reasoning and problem solving. The medial frontal and frontoparietal networks, which comprised brain regions within the frontal, parietal, and temporal lobes that are associated with cognitive control and altering connection patterns during tasks, were the most distinct between individuals and the strongest predictors of fluid intelligence.

This study highlights the possibility of discovering individual-level markers of brain connectivity and related behavior that could be used to personalize educational and clinical interventions.

DOI:10.1038/nn.4135 | Original article

Research highlights

PrivacyMark System