Spontaneous cortical activity and self-initiated movement


Event details

Date 10.10.2012
Hour 13:1514:15
Speaker Dr. Aaron Schurger, INCERM, FR.
Category Conferences - Seminars
The origin of voluntary, self-initiated movements is one of the most fascinating and important questions in neuroscience research. An influential finding in this area of research is that a gradual buildup of neuronal activity, known as the “readiness potential” (RP), reliably precedes self-initiated movements. In the early 1980’s, Benjamin Libet found that the onset of the RP precedes the conscious “urge” to move by 300 ms or more (Libet et al, 1983), and more recent study confirms the pre-urge buildup at the single-neuron level (Fried et al, 2011). A related experiment using fMRI showed that binary decisions could be predicted with better-than-chance accuracy, several seconds before the decision was reached (Soon et al, 2008). Experiments such as these have had an unrivaled influence on the prevailing view that movement is initiated pre-consciously and the conscious “decision” to move is grafted on after the fact—leaving many to doubt that we have conscious control over our actions. This view, however, rests on the assumption that neural activity that reliably precedes self-initiated or self-chosen actions reflects the unconscious initiation of those actions.
Our recent work (Schurger, Sitt, & Dehaene PNAS 2012) directly challenges this assumption with the assertion that the RP reflects ongoing spontaneous fluctuations in neural activity, and not, as has long been thought, a specific goal-directed process. We demonstrate that neural-accumulator models of decision-making can explain the kind of results cited above, once the autocorrelated nature of spontaneous brain activity is taken into account. Our theoretical model provides answers to puzzling questions that are not easily answered under the prevailing framework – for example, why do crayfish also have a readiness potential, and why is the RP not an exceptionally good predictor of movement onset? I will describe our model, and the evidence supporting it, and will discuss its implications for the prediction of behavior from brain activity, the development of asynchronous BCI's, and the role of brain-body interactions in the initiation of movement. I will propose that a central theme connecting these different domains is the question of when to move, and in closing I will briefly describe work-in-progress on the development of a short-latency MEG/EEG-based asynchronous BCI using the ADA-boost machine-learning algorithm. I will use this project to highlight the potential synergy between BCI research, self-initiated-movement research, and schizophrenia research.

Practical information

  • Informed public
  • Free


  • Center for Neuroprosthetics

Event broadcasted in