In the initial stage of the project, we have initiated two new positions and research programs. One PhD program on neurofeedback in fMRI/EEG and one on machine learning for discovering a coupling model between fMRI and EEG signal:

  • The PhD program is devoted to search for innovative paths on both sides, i.e., 1) the exploitation of a novel combination of several cerebral recording devices (real-time EEG *plus* fMRI devices), and 2) the exploitation of novel feedback strategies based on multi-sensory feedback and virtual reality technologies. More specifically, this PhD will study the design of novel paradigms of Neurofeedback under visual or auditory stimuli that can be conducted with fMRI coupled with simultaneous EEG recordings. The targeted pathologies are related to functional rehabilitation (stroke and ADHD) and psychiatric disorders (depressions). Evaluations will be conducted in close collaborations with the medical doctors partners of the HEMISFER project, especially Prof. I. BONAN (Visages U746, Rennes Hospital CHU) and Prof. D. DRAPIER (EA 4712, Rennes psychiatric hospital CHGR). At the present stage, this PhD addresses the definition of novel experimental paradigms of immersive stimulations for Neurofeedback based on fMRI and EEG. It will study the design of novel sensory feedbacks and novel immersive visualization techniques to display cerebral activity coming from fMRI and EEG.  Stereoscopic visual feedback (3D), but also auditory feedback will be proposed to enhance the immersion and motivation of the patient, and the perception of the evolution of his/her brain activity.
  • The aim of the post-doc is to design a new brain imaging procedure through the simultaneous acquisition of EEG and fMRI. By combining both modalities, one expects to achieve a good reconstruction both in time and space. This new imaging technique will then be used for improving neurofeedback paradigms in the context of rehabilitation and psychiatric disorders, which is the final purpose of the HEMISFER project. Numerous studies have been devoted to the simultaneous acquisition of EEG and fMRI data, and to their pre-processing. As detailed in [1], integration of fMRI with EEG has been already intensively pursued in two directions: spatial coupling, where the activity map provided by fMRI provides an additional a priori information [2]; and temporal coupling, where the EEG/MEG dynamic signatures in the time or frequency domain inform the statistical mapping of fMRI. We are more interested in a symmetric coupling, where the totality of each modality is used equally.  In this case, the key point is to model the link between the electrophysological information (EEG) and the hemodynamic measurements (BOLD signal measured through fMRI): this is referred as the neurovascular coupling. The simplest coupling involves a linear relation, where the BOLD response is obtained by convoluting the neuronal activity with a function h similar to the canonical hemodynamic response function (HRF).  A theoretical foundation can be found in [3], which linearizes the relationship between the neuronal activity and the BOLD signal according to the well known balloon model [4].


  1. Bin He, Lin Yang, Christopher Wilke, and Han Yuan. Electrophysiological imaging of brain activity and connectivity—challenges and opportunities. Biomedical Engineering, IEEE Transactions on, 58(7):1918–1931, 2011.
  2. Zhongming Liu and Bin He. fmri–eeg integrated cortical source imaging by use of timevariant spatial constraints. NeuroImage, 39(3):1198–1214, 2008.
  3. PA Robinson, PM Drysdale, Helena Van der Merwe, Elizabeth Kyriakou, MK Rigozzi, Biljana Germanoska, and CJ Rennie. Bold responses to stimuli: dependence on frequency, stimulus form, amplitude, and repetition rate. Neuroimage, 31(2):585–599, 2006.
  4. Karl J Friston, A Mechelli, R Turner, and CJ Price. Nonlinear responses in fmri: the balloon model, volterra kernels, and other hemodynamics. NeuroImage, 12(4):466–477, 2000.

Comments are closed.