Concurrent fNIRS/fMRI recordings represent multiple, simultaneously active, regionally overlapping neural hemodynamic responses. In this study, we propose a novel parallel framework for the spatiotemporal fNIRS/fMRI fusion to address the issues due to the overlapping nature of these responses. The developed fusion techniques employ the Independent Component Analysis (ICA) to recover the time courses and spatial mapping components from fNIRS and fMRI separately. Then the correlated components from each imaging modality are combined concurrently in the spatial and temporal domain for fMRI-guided fNIRS and fNIRS aided fMRI.