MAGNETOENCEPHALOGRAPHIC (MEG) INVESTIGATION OF PARALLEL PROCESSING IN THE HUMAN VISUAL BRAIN
Dr Stephen Anderson, Department of Psychology, Royal Holloway, University of London

Abstract: Electrical current flow in individuals produces magnetic fields outside the individual that can be measured using SQUID (superconducting quantum interference device) magnetometers. Early research concentrated on the human heart, but technological advances have enabled magnetic fields generated by brain activity to be recorded. This is a remarkable achievement as the strength of the magnetic field emerging from the human brain is about seven orders of magnitude smaller than that in a typical urban environment. The technique is called magneto-encephalography (MEG). There is a general consensus that the measured magnetic fields reflect extracellular ionic current flow, and as such relate to neural activity. This is a clear advantage over PET and fMRI, which only provide an indirect measure of neural activity as both methods rely upon associated metabolic events within the brain for functional imaging. Another major advantage of MEG is that brain activity can be resolved on a time scale that is compatible with human thought processes.

In this talk I will discuss how MEG can be used to investigate parallel processing in the human visual brain. To demonstrate this, I will concentrate on motion processing in the human visual system. Evidence for the existence of an area within the human brain specialised for the analysis of motion has come from behavioural studies on brain-damaged patients and most recently from PET and fMRI studies. This area is thought to be the human homologue of the primate cortical motion area V5 (MT). MEG places human V5 near the temporo-parieto-occipital junction. MEG's non-invasive nature and rapid data acquisition rate allowed the temporal dynamics and functional response properties of this area to be determined, and these details will be reported in the lecture. A second series of MEG experiments on a subject with extensive damage to the primary visual area (V1), allowed us to establish that there are at least two different anatomical routes by which information reaches V5, one via V1 and the other not. Details of this study will also be reported.

This seminar was held at the Department of Computer Science, Royal Holloway, University of London on 4 June 1997

back