Rofo 2008; 180 - A7
DOI: 10.1055/s-2008-1060345

Making EMG recordings during fMRI work: experiences from fundamental and applied studies of the motor system

NM Maurits 1, 2, RJ Renken 2, JM Hoogduin 2, H van Duinen 3, I Zijdewind 2, 3, MAJ Tijssen 4, AF van Rootselaar 4, 5
  • 1University Medical Center Groningen, University of Groningen, Department of Neurology, the Netherlands
  • 2BCN-NeuroImaging Center, University of Groningen, the Netherlands
  • 3University Medical Center Groningen, University of Groningen, Department of Medical Physiology, the Netherlands
  • 4Academic Medical Center Amsterdam, Department of Neurology, the Netherlands
  • 5Academic Medical Center Amsterdam, Department of Clinical Neurophysiology, the Netherlands

The (dys)functioning motor system provides a challenging system to study. For a thorough understanding of the human motor system, simultaneous measurements of as many involved subsystems as possible, would be a great advantage. In animal studies the relation between brain and muscle activity, has been studied with intracranial recordings together with EMG. In humans, incidental intracranial recordings are available from surgical patients, e.g. preceding epilepsy surgery or after implantation of electrodes for deep brain stimulation in Parkinson's disease. Additionally, combined EEG/MEG-EMG recordings can be applied. A drawback of these approaches is the limited spatial coverage and/or resolution. A non-invasive technique that could provide information on entire brain and muscle activity at the same time, was not available for a long time. Almost three years ago, we started recording EMG and fMRI simultaneously in Groningen, thinking that these recordings of the entire brain ánd muscle activity would allow us to gain more understanding of the human motor system.

Not surprisingly, from first recordings to first (clinical) applications has proven to be quite a challenge. Even though (truly) simultaneous EEG-fMRI recordings have been performed successfully since the late 1990's, and EMG is technically not much different from EEG, EMG recordings during fMRI provide new problems to be solved, such as the presence of movement artefacts that can hamper conventional artefact correction methods. Here, we focus on the path towards these applications. We discuss different technical approaches to recording EMG during fMRI (e.g. monopolar vs. bipolar derivations) and show some of the details involved in optimizing the EMG and fMRI recordings with respect to each other (such as reference and electrode wire positions). Further we demonstrate how optimal paradigms, designs and analyses can be chosen for particular experiments. Finally we illustrate problems and solutions in correcting EMG recordings for fMRI and movement artefacts and discuss different ways to integrate EMG and fMRI analyses.

To conclude, we will show some of the results in patients, with an outlook to the future of this new and exciting technique.