Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore a novel Julia implementation of spectral Dynamic Causal Modeling (DCM) that significantly enhances computational performance and modeling flexibility for neural connectivity inference. Learn how this 11-minute conference talk demonstrates leveraging Julia's automatic differentiation capabilities to achieve up to tenfold speed improvements over traditional MATLAB-based SPM implementations while maintaining parameter estimation accuracy. Discover how the implementation utilizes Julia's ModelingToolkit for modular model assembly, enabling more flexible modeling environments compared to existing solutions. Understand the validation process using both simulated and empirical fMRI data, and examine how these computational improvements make DCM applicable to larger brain region networks by reducing the practical limitations imposed by computational cost. Gain insights into the open-source platform's potential for advancing neural connectivity research and parameter fitting in neuroscience, with detailed discussion of the biophysically detailed models used in spectral domain parameter estimation and the variational Bayes estimation techniques underlying the DCM framework.
