Concurrent acquisition of functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) is widely used to monitor the neuronal activities of brain. However, this simultaneous recording suffers from complex artifacts. The Ballistocardiogram (BCG) artifact in specific, is as yet poorly assumed, appears to be more challenging and hinders to exploit the full strength of both modalities. In this paper, a hybrid method is implemented which combines Empirical Mode Decomposition (EMD) with Adaptive Filtering (AF) using notch filter to reduce the BCG artifact. Results of this study demonstrate that the proposed algorithm is generally useful and effective for the reduction of the BCG artifact. © Springer-Verlag 2013.