Sounds evoked activity in a large fraction of V1 neurons, and this activity was reliably different across sounds. j– l, Same as g– i, for individual mice (thin curves) and averaged across mice (thick curves). h, Fraction of total variance explained by auditory PCs, for this example mouse inset: distribution of the weights of auditory PC1 (arbitrary units), showing that weights were typically positive. g, Time courses of the first principal component of the sound-related responses in e (‘auditory PC1’, arbitrary units). f, Decoding accuracy for video versus sound (double asterisks indicate P = 0.0039, Wilcoxon right-tailed signed rank test, n = 8 mice). e, Sound-related time courses for all 212 neurons in one experiment, sorted using rastermap 13. Scale bars in b– d: 20 spikes per second. d, Grand average over all conditions for the neuron. c, Same, for the sound-related time courses. b, Video-related time courses (averaged over all sound conditions, minus the grand average) for the example neuron in a. Responses were averaged over four repeats. Sounds evoke stereotyped responses in visual cortex.Ī, Responses of an example neuron to combinations of sounds (columns) and videos (rows). Thus much of the multisensory activity that has been observed in visual cortex may have a simpler, behavioral origin. These movements were small but specific to each sound and stereotyped across trials and across mice. Furthermore, it was independent of direct projections from auditory cortex, and it tightly correlated with the uninstructed movements evoked by the sounds. Moreover, it was essentially identical to activity evoked in another brain region, the hippocampal formation. As predicted by our hypothesis, the activity evoked by sounds in V1 had a low dimension: it was largely one-dimensional. To test these predictions, we recorded the responses of hundreds of neurons in mouse V1 to audiovisual stimuli, while filming the mouse to assess the movements elicited by the sounds. Moreover, sound-evoked activity should be independent of direct inputs from auditory cortex and should be predictable from the behavioral effects of sounds. This hypothesis predicts that sound-evoked activity in V1 should have the typical attributes of behavioral signals: low dimension 13 and a broad footprint 14, 21, 22 that extends beyond the cortex 13. This seems possible because sounds can change internal state and evoke uninstructed body movements 16– 20. We hypothesized, therefore, that the activity evoked by sounds in V1 reflects sound-elicited changes in internal state and behavior. These behavioral and state signals are low-dimensional and largely orthogonal 13 to the high-dimensional code that V1 uses to represent images 15. For instance, the activity of V1 neurons carries strong signals related to running 10, 11, pupil dilation 11, 12, whisking 13 and other movements 14. Behavioral and state signals have profound effects on sensory areas. Here we consider a possible alternative explanation for these multisensory signals, based on low-dimensional changes in internal state and behavior 8, 9. This sound-evoked activity is thought to originate from direct projections from the auditory cortex 2, 3, 5, 7-it may be suppressed by inhibition of the auditory cortex 2, 5, and it may be mimicked by stimulation of auditory fibers 2, 3. Sounds may provide V1 with global inhibition 2, modify the neurons’ tuning 3, 4, boost detection of visual events 5 or even provide tone-specific information, reinforced by prolonged exposure 6 or training 7. For instance, mouse’s primary visual cortex (V1) is influenced by sounds. Many studies suggest that all cortical sensory areas, including primary ones, are multisensory 1.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |