Linguistic representation of vowels in speech imagery EEG.

Speech imagery recognition from electroencephalograms (EEGs) could potentially become a strong contender among non-invasive brain-computer interfaces (BCIs). In this report, first we extract language representations as the difference of line-spectra of phones by statistically analyzing many EEG signals from the Broca area. Then we extract vowels by using iterative search from hand-labeled short-syllable data. The iterative search process consists of principal component analysis (PCA) that visualizes linguistic representation of vowels through eigen-vectors φ(m), and subspace method (SM) that searches an optimum line-spectrum for redesigning φ(m). The extracted linguistic representation of Japanese vowels /i/ /e/ /a/ /o/ /u/ shows 2 distinguished spectral peaks (P1, P2) in the upper frequency range. The 5 vowels are aligned on the P1-P2 chart. A 5-vowel recognition experiment using a data set of 5 subjects and a convolutional neural network (CNN) classifier gave a mean accuracy rate of 72.6%.

Cortical Activation Patterns Evoked by Temporally Asymmetric Sounds and Their Modulation by Learning.

When complex sounds are reversed in time, the original and reversed versions are perceived differently in spectral and temporal dimensions despite their identical duration and long-term spectrum-power profiles. Spatiotemporal activation patterns evoked by temporally asymmetric sound pairs demonstrate how the temporal envelope determines the readout of the spectrum. We examined the patterns of activation evoked by a temporally asymmetric sound pair in the primary auditory field (AI) of anesthetized guinea pigs and determined how discrimination training modified these patterns. Optical imaging using a voltage-sensitive dye revealed that a forward ramped-down natural sound (F) consistently evoked much stronger responses than its time-reversed, ramped-up counterpart (revF). The spatiotemporal maximum peak (maxP) of F-evoked activation was always greater than that of revF-evoked activation, and these maxPs were significantly separated within the AI. Although discrimination training did not affect the absolute magnitude of these maxPs, the revF-to-F ratio of the activation peaks calculated at the location where hemispheres were maximally activated (i.e., F-evoked maxP) was significantly smaller in the trained group. The F-evoked activation propagated across the AI along the temporal axis to the ventroanterior belt field (VA), with the local activation peak within the VA being significantly larger in the trained than in the naïve group. These results suggest that the innate network is more responsive to natural sounds of ramped-down envelopes than their time-reversed, unnatural sounds. The VA belt field activation might play an important role in emotional learning of sounds through its connections with amygdala.

Auditory-visual integration in fields of the auditory cortex.

While multimodal interactions have been known to exist in the early sensory cortices, the response properties and spatiotemporal organization of these interactions are poorly understood. To elucidate the characteristics of multimodal sensory interactions in the cerebral cortex, neuronal responses to visual stimuli with or without auditory stimuli were investigated in core and belt fields of guinea pig auditory cortex using real-time optical imaging with a voltage-sensitive dye. On average, visual responses consisted of short excitation followed by long inhibition. Although visual responses were observed in core and belt fields, there were regional and temporal differences in responses. The most salient visual responses were observed in the caudal belt fields, especially posterior (P) and dorsocaudal belt (DCB) fields. Visual responses emerged first in fields P and DCB and then spread rostroventrally to core and ventrocaudal belt (VCB) fields. Absolute values of positive and negative peak amplitudes of visual responses were both larger in fields P and DCB than in core and VCB fields. When combined visual and auditory stimuli were applied, fields P and DCB were more inhibited than core and VCB fields beginning approximately 110 ms after stimuli. Correspondingly, differences between responses to auditory stimuli alone and combined audiovisual stimuli became larger in fields P and DCB than in core and VCB fields after approximately 110 ms after stimuli. These data indicate that visual influences are most salient in fields P and DCB, which manifest mainly as inhibition, and that they enhance differences in auditory responses among fields.

Recognition of Modified Conditioning Sounds by Competitively Trained Guinea Pigs.

The guinea pig (GP) is an often-used species in hearing research. However, behavioral studies are rare, especially in the context of sound recognition, because of difficulties in training these animals. We examined sound recognition in a social competitive setting in order to examine whether this setting could be used as an easy model. Two starved GPs were placed in the same training arena and compelled to compete for food after hearing a conditioning sound (CS), which was a repeat of almost identical sound segments. Through a 2-week intensive training, animals were trained to demonstrate a set of distinct behaviors solely to the CS. Then, each of them was subjected to generalization tests for recognition of sounds that had been modified from the CS in spectral, fine temporal and tempo (i.e., intersegment interval, ISI) dimensions. Results showed that they discriminated between the CS and band-rejected test sounds but had no preference for a particular frequency range for the recognition. In contrast, sounds modified in the fine temporal domain were largely perceived to be in the same category as the CS, except for the test sound generated by fully reversing the CS in time. Animals also discriminated sounds played at different tempos. Test sounds with ISIs shorter than that of the multi-segment CS were discriminated from the CS, while test sounds with ISIs longer than that of the CS segments were not. For the shorter ISIs, most animals initiated apparently positive food-access behavior as they did in response to the CS, but discontinued it during the sound-on period probably because of later recognition of tempo. Interestingly, the population range and mean of the delay time before animals initiated the food-access behavior were very similar among different ISI test sounds. This study, for the first time, demonstrates a wide aspect of sound discrimination abilities of the GP and will provide a way to examine tempo perception mechanisms using this animal species.

Discrete-time quantum walk with feed-forward quantum coin.

Constructing a discrete model like a cellular automaton is a powerful method for understanding various dynamical systems. However, the relationship between the discrete model and its continuous analogue is, in general, nontrivial. As a quantum-mechanical cellular automaton, a discrete-time quantum walk is defined to include various quantum dynamical behavior. Here we generalize a discrete-time quantum walk on a line into the feed-forward quantum coin model, which depends on the coin state of the previous step. We show that our proposed model has an anomalous slow diffusion characterized by the porous-medium equation, while the conventional discrete-time quantum walk model shows ballistic transport.

Recognition of non-harmonic natural sounds by small mammals using competitive training.

Animals recognize biologically relevant sounds, such as the non-harmonic sounds made by some predators, and respond with adaptive behaviors, such as escaping. To clarify which acoustic parameters are used for identifying non-harmonic, noise-like, broadband sounds, guinea pigs were conditioned to a natural target sound by introducing a novel training procedure in which 2 or 3 guinea pigs in a group competed for food. A set of distinct behavioral reactions was reliably induced almost exclusively to the target sound in a 2-week operant training. When fully conditioned, individual animals were separately tested for recognition of a set of target-like sounds that had been modified from the target sound, with spectral ranges eliminated or with fine or coarse temporal structures altered. The results show that guinea pigs are able to identify the noise-like non-harmonic natural sounds by relying on gross spectral compositions and/or fine temporal structures, just as birds are thought to do in the recognition of harmonic birdsongs. These findings are discussed with regard to similarities and dissimilarities to harmonic sound recognition. The results suggest that similar but not identical processing that requires different time scales might be used to recognize harmonic and non-harmonic sounds, at least in small mammals.

Spatiotemporal dynamics of neural activity related to auditory induction in the core and belt fields of guinea-pig auditory cortex.

Auditory induction is a continuity illusion in which missing sounds are perceived under appropriate conditions, for example, when noise is inserted during silent gaps in the sound. To elucidate the neural mechanisms underlying auditory induction, neural responses to tones interrupted by a silent gap or noise were examined in the core and belt fields of the auditory cortex using real-time optical imaging with a voltage-sensitive dye. Tone stimuli interrupted by a silent gap elicited responses to the second tone following the gap as well as early phasic responses to the first tone. Tone stimuli interrupted by broad-band noise (BN), considered to cause auditory induction, considerably reduced or eliminated responses to the tone following the noise. This reduction was stronger in the dorsocaudal field (field DC) and belt fields compared with the anterior field (the primary auditory cortex of guinea pig). Tone stimuli interrupted by notched (band-stopped) noise centered at the tone frequency, considered to decrease the strength of auditory induction, partially restored the second responses from the suppression caused by BN. These results suggest that substantial changes between responses to silent gap-inserted tones and those to BN-inserted tones emerged in field DC and belt fields. Moreover, the findings indicate that field DC is the first area in which these changes emerge, suggesting that it may be an important region for auditory induction of simple sounds.

Dynamic spatiotemporal inhibition in the guinea pig auditory cortex.

Real-time optical imaging was conducted in the guinea pig auditory cortex to study spatiotemporal interrelations of excitation and inhibition in response to tone stimulation. Tone stimulation elicited responses consisting of three phases in the anterior field (the primary auditory cortex of guinea pig) and in the dorsocaudal field of the auditory cortex. An early depolarization was followed by a late hyperpolarization and an even later depolarization both in the maximum excitatory regions and in the lateral regions beside and/or between them. The late hyperpolarization began significantly earlier and was stronger in the lateral regions than in the maximum excitatory regions. These results show that inhibition is dynamic, both in time and in space, in the auditory cortex.

Layer-specific short-term dynamics in network activity in the cerebral cortex.

Optical imaging with a voltage-sensitive dye was conducted in frontal slices of rat auditory cortex to study spatiotemporal patterns of response to repetitive electrical stimulation. When the rate of repetitive stimulation increased to 40 Hz, the amplitude ratio of the response after the fifth stimulus to the response after the first stimulus was significantly smaller in layers II/III than in layer IV or in layers V/VI. Similar results were obtained regardless of where electrical stimulation was applied. When the rate of stimulation was at 10 Hz, no difference was observed in response ratios among layers. These results show that the manner of responding to repetitive stimulation is different among layers and that it is stimulation rate-dependent.

Optical imaging of binaural interaction in multiple fields of the guinea pig auditory cortex.

Locating the source of a sound is an important function of the auditory system and interaural intensity differences are one of the most important cues. To study the functional pathways of sound localisation processing in the auditory cortex, activity in multiple fields of the guinea pig auditory cortex during stimulation with interaural intensity differences was studied using optical imaging with a voltage-sensitive dye. Of the auditory core (primary and dorsocaudal) and the belt fields which surround them, the posterior and ventroposterior belt fields were the most sensitive to interaural intensity differences. This suggests that the caudal pathway of the auditory cortex is involved in sound localisation.

Spatial focusing of neuronal responses induced by asynchronous two-tone stimuli in the guinea pig auditory cortex.

Spatiotemporal patterns of neuronal responses to asynchronous two-tone stimuli in the anterior field of the auditory cortex of anesthetized guinea pigs were studied using an optical recording method (12 x 12 photodiode array, voltage sensitive dye RH795). Interactions between the onset response to the first tone (masker; 5, 8, 10, 12 and 15 kHz, 200 ms) and to the second tone (probe; 10 kHz, 30 ms) with onset delays relative to the masker onset (0, 5, 10, 15 and 20 ms) were investigated. In general, two-tone interaction was suppressive rather than facilitative. At 0-10 ms probe delays, two-tone responses induced in the probe isofrequency area on the cortex tended to fuse with the masker response. At 15-20 ms probe delays, the probe response was apparently reduced, but was spatially focused and separated from the masker response. This spatial focusing of the probe response may have been due to neuronal inhibition originating after the masker onset response. These results are in agreement with psychoacoustical observations in human subjects, such as auditory segregation, and indicate that the spatial focusing of the cortical response provides a neuronal basis for detecting slightly asynchronous auditory inputs.