Timing of V1/V2 and V5+ activations during coherent motion of dots: an MEG study

In order to study the temporal activation course of visual areas V1 and V5 in response to a motion stimulus, a random dots kinematogram paradigm was applied to eight subjects while magnetic fields were recorded using magnetoencephalography (MEG). Sources generating the registered magnetic fields were localized with Magnetic Field Tomography (MFT). Anatomical identification of cytoarchitectonically defined areas V1/V2 and V5 was achieved by means of probabilistic cytoarchitectonic maps. We found that the areas V1/V2 and V5+ (V5 and other adjacent motion sensitive areas) exhibited two main activations peaks at 100-130 ms and at 140-200 ms after motion onset. The first peak found for V1/V2, which corresponds to the visual evoked field (VEF) M1, always preceded the peak found in V5+. Additionally, the V5+ peak was correlated significantly and positively with the second V1/V2 peak. This result supports the idea that the M1 component is generated not only by the visual area V1/V2 (as it is usually proposed), but also by V5+. It reflects a forward connection between both structures, and a feedback projection to V1/V2, which provokes a second activation in V1/V2 around 200 ms. This second V1/V2 activation (corresponding to motion VEF M2) appeared earlier than the second V5+ activation but both peaked simultaneously. This result supports the hypothesis that both areas also generate the M2 component, which reflects a feedback input from V5+ to V1/V2 and a crosstalk between both structures. Our study indicates that during visual motion analysis, V1/V2 and V5+ are activated repeatedly through forward and feedback connections and both contribute to m-VEFs M1 and M2.