Yida Wang1, Jonathan D Cohen2, Kai Li3, Nicholas B Turk-Browne4. 1. Department of Computer Science, Princeton University, Princeton, NJ, 08544, United States. Electronic address: yidawang@princeton.edu. 2. Department of Psychology, Princeton University, Princeton, NJ, 08544, United States; Princeton Neuroscience Institute, Princeton University, Princeton, NJ, 08544, United States. Electronic address: jdc@princeton.edu. 3. Department of Computer Science, Princeton University, Princeton, NJ, 08544, United States. Electronic address: li@cs.princeton.edu. 4. Department of Psychology, Princeton University, Princeton, NJ, 08544, United States; Princeton Neuroscience Institute, Princeton University, Princeton, NJ, 08544, United States. Electronic address: ntb@princeton.edu.
Abstract
BACKGROUND: The analysis of brain imaging data often requires simplifying assumptions because exhaustive analyses are computationally intractable. Standard univariate and multivariate analyses of brain activity ignore interactions between regions and analyses of interactions (functional connectivity) reduce the computational challenge by using seed regions of interest or brain parcellations. NEW METHOD: To meet this challenge, we developed full correlation matrix analysis (FCMA), which leverages and optimizes algorithms from parallel computing and machine learning to efficiently analyze the pairwise correlations of all voxels in the brain during different cognitive tasks, with the goal of identifying task-related interactions in an unbiased manner. RESULTS: When applied to a localizer dataset on a small compute cluster, FCMA accelerated a naive, serial approach by four orders of magnitude, reducing running time from two years to one hour. In addition to this performance gain, FCMA emphasized different brain areas than existing methods. In particular, beyond replicating known category selectivity in visual cortex, FCMA also revealed a region of medial prefrontal cortex whose selectivity derived from differential patterns of functional connectivity across categories. COMPARISON WITH EXISTING METHOD(S): For benchmarking, we started with a naive approach and progressively built up to the complete FCMA procedure by adding optimized classifier algorithms, multi-threaded parallelism, and multi-node parallelism. To evaluate what can be learned with FCMA, we compared it against multivariate pattern analysis of activity and seed-based analysis of functional connectivity. CONCLUSIONS: FCMA demonstrates how advances in computer science can alleviate computational bottlenecks in neuroscience. We have released a software toolbox to help others evaluate FCMA.
BACKGROUND: The analysis of brain imaging data often requires simplifying assumptions because exhaustive analyses are computationally intractable. Standard univariate and multivariate analyses of brain activity ignore interactions between regions and analyses of interactions (functional connectivity) reduce the computational challenge by using seed regions of interest or brain parcellations. NEW METHOD: To meet this challenge, we developed full correlation matrix analysis (FCMA), which leverages and optimizes algorithms from parallel computing and machine learning to efficiently analyze the pairwise correlations of all voxels in the brain during different cognitive tasks, with the goal of identifying task-related interactions in an unbiased manner. RESULTS: When applied to a localizer dataset on a small compute cluster, FCMA accelerated a naive, serial approach by four orders of magnitude, reducing running time from two years to one hour. In addition to this performance gain, FCMA emphasized different brain areas than existing methods. In particular, beyond replicating known category selectivity in visual cortex, FCMA also revealed a region of medial prefrontal cortex whose selectivity derived from differential patterns of functional connectivity across categories. COMPARISON WITH EXISTING METHOD(S): For benchmarking, we started with a naive approach and progressively built up to the complete FCMA procedure by adding optimized classifier algorithms, multi-threaded parallelism, and multi-node parallelism. To evaluate what can be learned with FCMA, we compared it against multivariate pattern analysis of activity and seed-based analysis of functional connectivity. CONCLUSIONS:FCMA demonstrates how advances in computer science can alleviate computational bottlenecks in neuroscience. We have released a software toolbox to help others evaluate FCMA.
Authors: Judith E Fan; Jeffrey D Wammes; Jordan B Gunn; Daniel L K Yamins; Kenneth A Norman; Nicholas B Turk-Browne Journal: J Neurosci Date: 2019-12-23 Impact factor: 6.167
Authors: Jiahe Zhang; Joseph M Andreano; Bradford C Dickerson; Alexandra Touroutoglou; Lisa Feldman Barrett Journal: Cereb Cortex Date: 2020-01-10 Impact factor: 5.357
Authors: Jonathan D Cohen; Nathaniel Daw; Barbara Engelhardt; Uri Hasson; Kai Li; Yael Niv; Kenneth A Norman; Jonathan Pillow; Peter J Ramadge; Nicholas B Turk-Browne; Theodore L Willke Journal: Nat Neurosci Date: 2017-02-23 Impact factor: 24.884
Authors: Marcelo G Mattar; Michael W Cole; Sharon L Thompson-Schill; Danielle S Bassett Journal: PLoS Comput Biol Date: 2015-12-02 Impact factor: 4.475
Authors: Vikranth R Bejjanki; Rava Azeredo da Silveira; Jonathan D Cohen; Nicholas B Turk-Browne Journal: PLoS Comput Biol Date: 2017-08-25 Impact factor: 4.475