hide
Free keywords:
-
Abstract:
The local field potential (LFP) comprises the low frequency membrane potential fluctuations
of an extracellular recording. It is thought to be closely related to frequently used, noninvasively
recorded signals such as the EEG or the BOLD signal and has been proven to show
some stimulus selectivity in various brain regions. Thus the relationship of the LFP to the
spiking activity as well as a comparison of their level of selectivity is of great interest. We
simultaneously recorded spiking activity and LFPs from tree shrew primary visual cortex
using pairs or triplets of tetrodes separated by between 200 and 1200μm. We map the receptive
fields at eccentricities between 10 and 25deg using binary sparse noise and estimate
their spatial extent by fitting oriented two dimensional Gaussians to the resulting activation
map. We have preliminary data from 49 pairs of simultaneously recorded neurons as well as
18 additional single sites from a total of 14 animals. The visual spread of the receptive field
estimated from the trial averaged LFP ( 2.99deg2) was statistically similar to the one for the
spiking activity ( 3.06deg2). Estimating the receptive field size using LFP gamma power (30-
90Hz), we found significantly smaller values ( 2.17deg2) compared to the above two conditions
(1-way ANOVA: p«0.001). In addition, we found that receptive fields tended to be ellipsoidal:
long to short axis ratio: 1.41 (LFP) and 1.36 (spikes) and oriented mostly horizontally: mean
angle 5.79deg (LFP) and 6.95deg (spikes). This bias toward horizontal orientations was statistically
significant (Rayleigh tests: p « 0.001) for both signal types. Relating the spatial
separation between the tetrodes to the respective receptive field centers, we found that 1deg
of visual angle corresponded to a cortical distance of around 180μm for both spikes and LFP.
Our results suggest that the spatial extent of activation estimated from LFPs can be similar
or smaller than the values for spiking activity depending on which features of the LFP
are analyzed. Thus, we obtained smaller receptive field sizes using gamma-band oscillations
compared to trial-averaged stimulus evoked LFPs. Our findings represent the first detailed
investigations of this relationship in tree shrew V1, and are in good general agreement with
related work in macaque monkeys.