Mark I. Sanderson

Time-frequency model for echo-delay resolution in wideband biosonar

A time/frequency model of the bats auditory system was developed to examine the basis for the fine (approximately 2 micros) echo-delay resolution of big brown bats (Eptesicus fuscus), and its performance at resolving closely spaced FM sonar echoes in the bats 20-100-kHz band at different signal-to-noise ratios was computed. The model uses parallel bandpass filters spaced over this band to generate envelopes that individually can have much lower bandwidth than the bats ultrasonic sonar sounds and still achieve fine delay resolution. Because fine delay separations are inside the integration time of the models filters (approximately 250-300 micros), resolving them means using interference patterns along the frequency dimension (spectral peaks and notches). The low bandwidth content of the filter outputs is suitable for relay of information to higher auditory areas that have intrinsically poor temporal response properties. If implemented in fully parallel analog-digital hardware, the model is computationally extremely efficient and would improve resolution in military and industrial sonar receivers.

Representation of waveform periodicity in the auditory midbrain of the bullfrog, Rana catesbeiana.

AbstractThe period of complex signals is encoded in the bullfrog’s eighth nerve by a synchrony code based on phase-locked responding. We examined how these arrays of phase-locked activity are represented in different subnuclei of the auditory midbrain, the torus semicircularis (TS). Recording sites in different areas of the TS differ in their ability to synchronize to the envelope of complex stimuli, and these differences in synchronous activity are related to response latency. Cells in the caudal principal nucleus (cell sparse zone) have longer latencies, and show little or no phase-locked activity, even in response to low modulation rates, while some cells in lateral areas of the TS (magnocellular nucleus, lateral part of principal nucleus) synchronize to rates as high as 90–100 Hz. At midlevels of the TS, there is a lateral-to-medial gradient of synchronization ability: cells located more laterally show better phase-locking than those located more medially. Pooled all-order interval histograms from short latency cells located in the lateral TS represent the waveform periodicity of a biologically relevant complex harmonic signal at different stimulus levels, and in a manner consistent with behavioral data from vocalizing male frogs. Long latency cells in the caudal parts of the TS (cell sparse zone, caudal magnocellular nucleus) code stimulus period by changes in spike rate, rather than by changes in synchronized activity. These data suggest that neural codes based on rate processing and time domain processing are represented in anatomically different areas of the TS. They further show that a population-based analysis can increase the precision with which temporal features are represented in the central auditory system.

Target representation of naturalistic echolocation sequences in single unit responses from the inferior colliculus of big brown bats

Echolocating big brown bats (Eptesicus fuscus) emit trains of frequency-modulated (FM) biosonar signals whose duration, repetition rate, and sweep structure change systematically during interception of prey. When stimulated with a 2.5-s sequence of 54 FM pulse-echo pairs that mimic sounds received during search, approach, and terminal stages of pursuit, single neurons (N = 116) in the bats inferior colliculus (IC) register the occurrence of a pulse or echo with an average of < 1 spike/sound. Individual IC neurons typically respond to only a segment of the search or approach stage of pursuit, with fewer neurons persisting to respond in the terminal stage. Composite peristimulus-time-histogram plots of responses assembled across the whole recorded population of IC neurons depict the delay of echoes and, hence, the existence and distance of the simulated biosonar target, entirely as on-response latencies distributed across time. Correlated changes in pulse duration, repetition rate, and pulse or echo amplitude do modulate the strength of responses (probability of the single spike actually occurring for each sound), but registration of the target itself remains confined exclusively to the latencies of single spikes across cells. Modeling of echo processing in FM biosonar should emphasize spike-time algorithms to explain the content of biosonar images.

Evaluation of an auditory model for echo delay accuracy in wideband biosonar.

In a psychophysical task with echoes that jitter in delay, big brown bats can detect changes as small as 10-20 ns at an echo signal-to-noise ratio of approximately 49 dB and 40 ns at approximately 36 dB. This performance is possible to achieve with ideal coherent processing of the wideband echoes, but it is widely assumed that the bats peripheral auditory system is incapable of encoding signal waveforms to represent delay with the requisite precision or phase at ultrasonic frequencies. This assumption was examined by modeling inner-ear transduction with a bank of parallel bandpass filters followed by low-pass smoothing. Several versions of the filterbank model were tested to learn how the smoothing filters, which are the most critical parameter for controlling the coherence of the representation, affect replication of the bats performance. When tested at a signal-to-noise ratio of 36 dB, the model achieved a delay acuity of 83 ns using a second-order smoothing filter with a cutoff frequency of 8 kHz. The same model achieved a delay acuity of 17 ns when tested with a signal-to-noise ratio of 50 dB. Jitter detection thresholds were an order of magnitude worse than the bat for fifth-order smoothing or for lower cutoff frequencies. Most surprising is that effectively coherent reception is possible with filter cutoff frequencies well below any of the ultrasonic frequencies contained in the bats sonar sounds. The results suggest that only a modest rise in the frequency response of smoothing in the bats inner ear can confer full phase sensitivity on subsequent processing and account for the bats fine acuity or delay.

Joint decoding of visual stimuli by IT neurons’ spike counts is not improved by simultaneous recording

Information about visual stimuli such as objects and faces is represented across populations of neurons of the inferior temporal cortex. Does recording from inferotemporal neurons simultaneously tell you more than recording from them sequentially? Equivalently, are neurons conditionally independent given a stimulus? To evaluate these issues, we recorded from two monkeys during a passive viewing task. Multiple neurons were simultaneously recorded on separate electrodes. From spike counts in 50-ms windows, we computed the mutual information between counts and images for each neuron individually and jointly with other simultaneously recorded neurons. To determine the significance of these values, we shuffled the stimulus labels (to test if there was significant information) or shuffled responses across trials involving the same image (to see if there was synergistic coding). We recorded from 127 pairs of neurons where each neuron individually was visually responsive. Depending on the time window, we found up to ∼ 90% of these pairs showed significant information about the visual stimulus. Shuffling across trials failed to show evidence for synergistic coding. In summary, if you were given two of our neuronal responses and asked to guess the stimulus which produced them you could not, in principle, do better with two simultaneously recorded spike counts than with any two spike counts selected randomly from trials of the same type.

The emergence of temporal hyperacuity from widely tuned cell populations.

Typically, individual neural cells operate on a millisecond time scale yet behaviorally animals reveal sub-microsecond acuity. Our model resolves this huge discrepancy by using populations of many widely tuned cells to attain sub-microsecond resolution in a temporal discrimination task. An echolocating bat uses its auditory system to locate objects and it demonstrates remarkable temporal precision in psychophysical tasks. Auditory cells were simulated using realistic parameters and connected in three ascending layers with descending projections from auditory cortex. Coincidence detection of firing collicular cells at thalamus and subsequent integration of multiple inputs at cortex, produce an estimate of time represented as the mean of the active cortical population. Multiple estimates allow the model bat to use memory to recognize predictable change in stimuli values. The best performance is produced using cortical feedback and a computation of target time based on combining the current and previous estimates. Temporal hyperacuity is attained through population coding of physiologically realistic cells but depends on the inherent properties of the psychophysical task.

Phase sensitivity of auditory brain‐stem responses in echolocating big brown bats

Multiple behavioral experiments show that echolocating big brown bats perceive 180° phase shifts of ultrasonic (20–100 kHz) FM echoes as delay changes of ±15 μm. These bats represent FM sweeps as coherent auditory spectrograms in which low‐pass smoothing of half‐wave‐rectified hair‐cell excitation is the critical limiting parameter. Computational modeling of auditory spectrograms combined with Monte Carlo simulation of echo delay psychophysics reveals that coherence is preserved when the auditory low‐ pass smoothing cutoff is as low as 7–10 kHz, which is not nearly as high as the 20‐ to 50‐kHz ultrasonic frequencies that seem necessary intuitively. Local‐field‐potential recordings from the bat’s auditory brain‐stem are sensitive to the starting phase of tone‐bursts at frequencies up to 14 kHz, manifested as a change in LFP wave‐form shape with phase. For most sites this phase sensitivity exhibits a strong dependence on stimulus amplitude. Typically, tone‐bursts of 65–75 dB SPL evoke significant changes in...

Biosonar model for improved range accuracy in a noisy environment

Using the theory of optimal receivers the range accuracy of echolocating systems can be expressed as a function of receiver bandwidth and signal-to-noise ratio through the well-known Woodward equation. That equation however was developed in the limit of very high signal-to-noise ratios, and assumes that the correct peak of the crosscorrelation function is known a-priori. Echolocating animals such as dolphins and bats have developed a highly specialized receiver to optimize echolocation to different environments and conditions. In particular, they use a set of filters with different center frequencies but overlapping bands. We show that this structure can help in improving accuracy in the case of relatively low signal-to-noise ratios when the ambiguity in the choice of the main peak of the crosscorrelation function cannot be avoided.

Delay axis of impulse-response images in biosonar

Summary form only given. Theoretical and experimental work on echo-delay acuity in bat sonar reveals a unique, previously unknown feature of the bats internal representation of sonar targets. Big brown bats (Eptesicus fuscus) broadcast wideband, frequency-modulated (FM) biosonar sounds containing frequencies of 20-100 kHz and perceive objects from echoes of these sounds that return to their ears. Echolocating bats determine target distance, or range, from the delay of echoes. These biosonar sounds have high center frequencies (fc) and wide bandwidths (?f), which in principle could support very accurate determination of echo delay. Using the auditory system as a sonar receiver, big brown bats are able to achieve this potential. For echoes with relatively unrestricted bandwidths, they can detect changes in echo delay as small as 10 to 40 ns at high echo signal-to-noise ratios. To understand how the wide frequency span of echoes contributes to the bats delay images, we restricted the frequency content of echoes by removing progressively greater portions of the low-frequency or the high-frequency end of the broadcast spectrum. We have found that the bats delay accuracy deteriorates systematically in a manner that depends on the ratio of echo center frequency to bandwidth, which equals the value for Q, the width of the target impulse response in number of cycles. The decline is not consistently proportional to either the reciprocal of the bandwidth or the reciprocal of the center frequency alone, as would be expected from sonar theory. The bats internal representation of targets appears to consist of the target impulse response scaled by the number of cycles rather than by units of time. Use of a normalized time axis for impulse responses may facilitate comparison of images for the same target at different aspects and could contribute to the bats ability to rapidly identify targets from their shape using a small number of pings.

A filterbank model for echo delay estimation in biosonar

Summary form only given. In a jittering-echo task, big brown bats achieve an echo-delay acuity of 10-20 ns (40 ns at echo signal-to-noise ratio of 36 dB). To understand the origin of this unusually fine acuity, a filterbank model of transduction in the bats peripheral auditory system was developed and tested in Monte Carlo simulations to determine its delay accuracy and test its acuity in a jittering-echo paradigm at different signal-to-noise levels. Several versions of the filterbank model were tested to learn how the models smoothing filter affected performance. When tested at an echo signal-to-noise ratio of 36 dB, the best filterbank delay-estimation method had a jitter threshold of 75 nanoseconds. To achieve 15 nanoseconds, this method required an echo signal-to-noise ratio of 50 dB and the use of a 2nd order lowpass smoothing filter with a cutoff frequency of 8 kHz. Jitter thresholds for filterbank models with either a lower cutoff frequency or a higher order were more than an order of magnitude worse than the behavioral threshold. These results predict that the smoothing filter for echolocating bats may have a higher cutoff frequency and shallower slope than typical mammalian values.