Objectives: 

The aim of this work was to study the effect of attention on the response of thalamic cells to a visual stimulus and the role of the corticofugal feedback on those attentional influences.

Materials: 

We recorded neuronal activity at the lateral geniculate nucleus (LGN) of two awake behaving monkeys during a visual task. Animals were trained to signal, by a subsequent eye movement, a stimulus that was different within a group of four (three identical distractors, one target). One of the stimuli was always placed over the receptive field (RF) of the cell under test. We compared the response elicited by the presence of the target versus the distractor over the RF. This protocol was repeated during V1 blockade by repetitive transcraneal magnetic stimulation (rTMS) at low frequency (0.8Hz) for 4 minutes. Visually evoked spike trains were separated into two components: Spikes fired in bursts (defined as groups of 2 or more action potentials whose interspike intervals were <6ms, preceded by a non-firing period of at least 50 msec) and tonic spikes (defined as all other spikes).

Results: 

Spikes fired in bursts were significantly more numerous when the stimulus on the RF was the target rather than one of the distractors, (chi2 p<0.05). Furthermore the distribution of burst spikes was also different between conditions, appearing earlier and being more concentrated in time when the target is presented (Kolmogorov-Smirnov P <0.001). Blockade of corticofugal input by TMS decreased neuronal responses, with a more potent effect on spikes in bursts, effectively eliminating the difference between target and distractors (Kolmogorov-Smirnov P>0.5). Interestingly, with TMS the decrease in the number of burst correlated with an increase in the mistakes made by the animal in the task.

Conclusions: 

Thus, cortical afferents to the LGN appear to modulate thalamic "detection" of a relevant visual stimulus through a specific effect on spikes fired in bursts.