American Statistical Association
The explosive growth of the human neuroimaging literature has led to major advances in understanding of human brain function, but has also made aggregation and synthesis of neuroimaging findings increasingly difficult. To address this problem, we recently introduced a highly automated brain mapping framework called Neurosynth that uses text mining, meta-analysis and machine learning techniques to automatically synthesize much of the functional neuroimaging literature. In this talk, I review a program of research that uses this framework to address a range of theoretical and methodological problems in cognitive neuroscience and neurogenetics. I illustrate how Neurosynth can be used for quantitative reverse inference, large-scale structure-to-function mapping, open-ended "decoding" of novel brain images, and genome-wide gene-cognition atlas development. I conclude with a discussion of current limitations and future directions.
|Date:||Tuesday, September 16, 2014|
“Central Park” meeting area at the Child Study Center
One Park Avenue (between 32nd and 33rd Streets)
New York, New York