Impact of differences between real and predicted time series on GLM fMRI analysis

• Authors: MIKL Michal; MAREČEK Radek; HLUŠTÍK Petr; BRÁZDIL Milan
• Year of publication: 2006
• Type: Conference abstract
• MU Faculty or unit: Faculty of Medicine
• Description: In functional magnetic resonance imaging (fMRI) detection of activation is often realized using massively univariate statistical methods. The measured signal is modeled using convolution of stimulus time-course and hemodynamic response function (HRF). For accurate fitting the data to the model it is necessary to know both HRF and stimulus time-course. The aim of this work is to find how much the results are depended on inaccurate knowledge of stimulus time-course.

Related projects

• The internal organisation and neurobiological mechanisms of functional CNS systems under normal and pathological conditions.