Bayesian Neural Response Model
The model generalizes the classical GLM by reducing to it when a known task predictor is supplied. BNR is intended as a model-based alternative to intersubject correlation for naturalistic fMRI. The Bayesian neural response model jointly e…
1 sources - 5 claims
The model generalizes the classical GLM by reducing to it when a known task predictor is supplied. BNR is intended as a model-based alternative to intersubject correlation for naturalistic fMRI. The Bayesian neural response model jointly estimates a latent shared BOLD response and voxel activation from fMRI data. The model treats response estimation and activation mapping as a single joint inference problem. For naturalistic fMRI, the shared response function is estimated from the data instead of prespecified.