A single link to the first track to allow the export script to build the search page
  • Addiction, Drugs
  • Information from Lay-Language Summaries is Embargoed Until the Conclusion of the Scientific Presentation

    014—Cocaine: Neural Mechanisms of Addiction I

    Saturday, November 09, 2013, 1:00 pm - 4:15 pm

    14.08: Human brain entropy mapping using thousands of subjects and its application in a drug addiction study

    Location: 33C

    *Z. WANG, J. SUH, Y. LI, Z. SINGER, R. EHRMAN, A. V. HOLE, C. P. O'BRIEN, A. R. CHILDRESS;
    Psychiatry, Univ. of Pennsylvania, PHILADELPHIA, PA

    Abstract Body: Entropy reflects the degrees of organization (vs. chaos, or disorganization) in a system. In the human brain, low entropy (a high degree of organization) is needed to function normally. Characterizing brain entropy (BEN) may thus provide an important way to assess brain states and brain functions, as well as its alterations by disorders such as drug addiction. The purpose of this study was to characterize BEN using resting state fMRI (rsfMRI) in normal brain and to assess its usability for drug addiction study. Entropy was calculated at each voxel of the brain and the BEN maps were registered into the same image space. We first showed that entropy of a nonliving water phantom doesn’t differ that from the environment, both are remarkably higher than that in the living brain. This sharp entropy difference proves the idea of using entropy for life detection when NASA designed their first Mars rover as well as the later established Gaia theory. Then, using a large cohort (n=1049) of normal subjects, we for the first time created a 3-dimensional BEN map showing relatively higher entropy in white matter but lower entropy in neocortex, which may reflect the “higher” mental functions subserved by cortex. Further analysis showed that BEN is self-organized into 7 hierarchical communities that are consistent with known structural and functional brain parcellations, suggesting using BEN as a physiologically and functionally meaningful brain activity measure. To assess the feasibility of using BEN for probing drug addiction, we calculated BEN from 43 chronic cocaine users and found that higher resting BEN in orbito-frontal cortex, ventral striatum, dorsolateral prefrontal cortex, and inferior temporal cortex, was positively correlated with higher drug dependence level (indexed by ASI drug severity rating), suggesting a less degree of organization in these interconnected
    frontal and limbic areas is associated with greater severity of the disorder, which is consistent with previous neurobiological studies and neuroimaging studies. Higher BEN in the so-called default mode network including posterior cingulate cortex/precuneus and parietal cortex was also positively correlated with cocaine craving (by self-report at baseline) and risk taking (mean adjusted pumps in Balloon Analog Risk task), two addiction-relevant phenomena. To summarize, we have revealed, for the first time, the spatial distribution of BEN in normal brain using a large sample size, and provided the first evidence that higher BEN in cocaine-addicted individuals has potential clinical significance, as it is linked to severity of drug dependence, craving, and risk-taking.

    Lay Language Summary: Entropy reflects the degrees of organization (vs. chaos, or disorganization) in a system. Human brain is a complex dynamic system, which needs low entropy (a high degree of organization) to function normally. Mapping brain entropy (BEN) may thus provide an important way to assess brain states and brain functions as well as their alterations due to brain disorders. The current study was performed to investigate BEN as a potential brain status marker as well as its implications in clinical applications.

    We first developed a robust way for measuring regional BEN using functional MRI (fMRI) with a whole brain resolution. We hypothesized that this BEN measure would discriminate between signals with different known regularities and would discriminate between neuronal and non-neuronal dynamics given the known fact of that living organism including our brain has much higher organizations than the non-living object. Our experiment clearly proved this hypothesis.
    Our second goal was to establish BEN distributions in the normal brain using large sample size. Human brain has been shown to reach a critical state that the neuronal dynamics present both spatial and temporal self-organizations. Theoretical work has also suggested a hierarchical organization of temporal depth in the cortex. According to these findings, we hypothesized that BEN in the normal brain would show a structured regional specificity and hierarchical self-organizations. Using resting state fMRI (rsfMRI) data from a large number of normal subjects (n=1049), we created a 3-dimensional BEN map showing relatively higher entropy in white matter but lower entropy in neocortex, which may reflect the “higher” mental functions (such as cognitive control or executive function) subserved by the cortex. In accordance with the functional brain inhomogeneity, BEN map shows spatial distributions. To further study the regional specificity and the potential hierarchy, we used an auto-clustering method to find the regional brain communities showing consistent BEN across normal subjects. We found the entire brain is self-organized into 7 hierarchical communities based on entropy of regional brain activity as measured by rsfMRI. The 7 clusters are consistent with known structural and functional brain parcellations such as the widely repeated resting state functional networks. These findings are consistent with theoretical model-based dynamic brain studies. Together they suggest that human brain presents long-range temporal organizations, which proves our hypothesis about brain entropy distribution and retrospectively prove using BEN as a physiologically and functionally meaningful brain activity measure.
    Our third goal was to assess the clinical implications of BEN mapping. We applied BEN mapping to rsfMRI data from 43 cocaine addicted patients. We found that higher resting BEN in orbito-frontal cortex, ventral striatum, dorsolateral prefrontal cortex, and inferior temporal cortex, was positively correlated with higher drug dependence level (indexed by ASI drug severity rating), suggesting a less degree of organization in these interconnected frontal and limbic areas is associated with greater severity of the disorder, which is consistent with previous neurobiological studies and neuroimaging studies. Higher BEN in the so-called default mode network including posterior cingulate cortex/precuneus and parietal cortex was also positively correlated with cocaine craving (by self-report at baseline) and risk taking (mean adjusted pumps in Balloon Analog Risk task), two addiction-relevant phenomena.
    To summarize, we have developed a BEN mapping method, revealed the spatial distribution of BEN in normal brain using a large sample size, and provided the first evidence that higher BEN in cocaine-addicted individuals has potential clinical significance, as it is linked to severity of drug dependence, craving, and risk-taking.