We study how the location of sounds is represented and encoded in the auditory cortex of awake and behaving marmosets, through full free-field or close-field sound presentation, with electrophysiological techniques.
(Remington and Wang, 2018) Unlike visual signals, sound can reach the ears from any direction, and the ability to localize sounds from all directions is essential for survival in a natural environment. Previous studies have largely focused on the space in front of a subject that is also covered by vision and were often limited to measuring spatial tuning along the horizontal (azimuth) plane. As a result, we know relatively little about how the auditory cortex responds to sounds coming from spatial locations outside the frontal space where visual information is unavailable. By mapping single-neuron responses to the full spatial field in awake marmoset (Callithrix jacchus), an arboreal animal for which spatial processing is vital in its natural habitat, we show that spatial receptive fields in several auditory areas cover all spatial locations. Several complementary measures of spatial tuning showed that neurons were tuned to both frontal space and rear space (outside the coverage of vision), as well as the space above and below the horizontal plane. Together, these findings provide valuable new insights into the representation of all spatial locations by primate auditory cortex.
Figure 3. Diversity of spatial receptive fields in auditory cortex.
(Zhou and Wang, 2014) The present study investigated two questions regarding forward suppression in the primary auditory cortex and adjacent caudal field of awake marmoset monkeys. First, what is the relationship between the location of a masker and its effectiveness in inhibiting neural response to a probe? Second, does varying the location of a masker change the spectral profile of forward suppression? We found that a masker can inhibit a neuron's response to a probe located at a preferred location even when the masker is located at a non-preferred location of a neuron. This is especially so for neurons in the caudal field. Furthermore, we found that the strongest forward suppression is observed when a masker's frequency is close to the best frequency of a neuron, regardless of the location of the masker. These results reveal, for the first time, the stability of forward masking in cortical processing of multiple sounds presented from different locations. They suggest that forward suppression in the auditory cortex is spectrally specific and spatially broad with respect to the frequency and location of the masker, respectively.
Figure 3. Example of forward suppression in response of two A1 neurons. (A) The rate-azimuth (AZ) tuning function and 2D SRF of an primary auditory cortex (A1) neuron in response to BF tone. The dashed line indicates the spontaneous rate. (B) Responses of the neuron shown in (A) to a pair of masker-probe stimuli. The top row shows the locations of masker and probe used for ‘co-located’, ‘near’ and ‘far’ conditions, respectively. The middle row shows the dot-raster plots of responses for each spatial arrangement. Light and dark gray mark the durations of masker and probe, respectively. For comparison, the probe-alone responses are given on the top of each raster plot. The bottom row shows rate–frequency tunings of excitation (red) and forward suppression (black) for each spatial arrangement. All tuning curves are plotted as a function of masker frequency (see text for the methods of calculation). Masker and probe were both pure-tone stimuli, whose parameters were given on the plot. The dashed line indicates BF.
(Zhou and Wang, 2012) Sound localization in both humans and monkeys is tolerant to changes in sound levels. The underlying neural mechanism, however, is not well understood. This study reports the level dependence of individual neurons' spatial receptive fields (SRFs) in the primary auditory cortex (A1) and the adjacent caudal field in awake marmoset monkeys.
Fig. 5. Effects of sound level on the spatial selectivity of example cortical neurons. A and B: spatial tuning of two A1 neurons. C and D: spatial tuning of two CM/CL neurons. Shown are a raster plot (top), rate-AZ tuning curve (middle), and SRF (bottom) of a neuron in response to 100-ms Gaussian broadband noise played at two sound pressure levels (SPLs). Data in the raster plots are organized based on the AZ angles of the loudspeakers; asterisks mark those at EL 45°. The gray shaded region marks the duration of a stimulus, and the SPL used is indicated above each raster plot. The dashed line on a rate-AZ tuning curve indicates the spontaneous rate of a neuron. BA was calculated as the ratio between the SRF area above 62.4% peak rate (outlined in blue) and the total area covered by the speaker array.