Modeling Percentage Change of fMRI BOLD Signal and Reaction Time of a Dual Task with a Queuing Network Modeling Approach

To model the percentage change of blood oxygenation level-dependent (BOLD) signal and reaction time in a dual task, we propose a new mathematical modeling approach—a queuing network approach based on queuing network theory of human performance (Liu, 1996, 1997) and current discoveries in neuroimage studies. This approach includes a queuing network architecture representing the information processing in the brain and mathematical equations to quantify the reaction time, BOLD signal and its percentage signal change (PSC). Both reaction time and the percentage change of BOLD signal in an fMRI study of the dual task are successfully modeled with analytical solutions of the mathematical equations, which demonstrates its usefulness and parsimony in modeling the brain activation pattern and human performance simultaneously. Furthermore, the current modeling approach uniquely quantifies the queuing mechanism discovered by the fMRI study and also provides a coherent and quantitative linkage between the neural signals and behavioral data. Further extension and development of the current modeling approach are discussed