Behavioral Plasticity of Audiovisual Perception: Rapid Recalibration of Temporal Sensitivity but Not Perceptual Binding Following Adult-Onset Hearing Loss

Abstract

The ability to accurately integrate or bind stimuli from more than one sensory modality is highly dependent on the features of the stimuli, such as their intensity and relative timing. Previous studies have demonstrated that the ability to perceptually bind stimuli is impaired in various clinical conditions such as autism, dyslexia, schizophrenia, as well as aging. However, it remains unknown if adult-onset hearing loss, separate from aging, influences audiovisual temporal acuity. In the present study, rats were trained using appetitive operant conditioning to perform an audiovisual temporal order judgment (TOJ) task or synchrony judgment (SJ) task in order to investigate the nature and extent that audiovisual temporal acuity is affected by adult-onset hearing loss, with a specific focus on the time-course of perceptual changes following loud noise exposure. In our first series of experiments, we found that audiovisual temporal acuity in normal-hearing rats was influenced by sound intensity, such that when a quieter sound was presented, the rats were biased to perceive the audiovisual stimuli as asynchronous (SJ task), or as though the visual stimulus was presented first (TOJ task). Psychophysical testing demonstrated that noise-induced hearing loss did not alter the rats' temporal sensitivity 2-3 weeks post-noise exposure, despite rats showing an initial difficulty in differentiating the temporal order of audiovisual stimuli. Furthermore, consistent with normal-hearing rats, the timing at which the stimuli were perceived as simultaneous (i.e., the point of subjective simultaneity, PSS) remained sensitive to sound intensity following hearing loss. Contrary to the TOJ task, hearing loss resulted in persistent impairments in asynchrony detection during the SJ task, such that a greater proportion of trials were now perceived as synchronous. Moreover, psychophysical testing found that noise-exposed rats had altered audiovisual synchrony perception, consistent with impaired audiovisual perceptual binding (e.g., an increase in the temporal window of integration on the right side of simultaneity; right temporal binding window (TBW)). Ultimately, our collective results show for the first time that adult-onset hearing loss leads to behavioral plasticity of audiovisual perception, characterized by a rapid recalibration of temporal sensitivity but a persistent impairment in the perceptual binding of audiovisual stimuli.

Keywords: audiovisual perception; hearing loss; multisensory processing; noise exposure; rat; synchrony judgment; temporal order judgment; temporal recalibration.

Figure 1

Rat audiovisual behavioral tasks and chamber set up. Rats were trained on either an audiovisual temporal order judgment (TOJ) task or a synchrony judgment (SJ) task. (A) Overview of both behavioral tasks. Through a series of stages, rats were trained using a two-alternative forced choice paradigm, where they were required to choose the right or left feeder trough depending on the stimulus condition presented. For example in the TOJ task, rats were trained to discriminate between auditory-first and visual-first trials, where the rats respond to the left feeder trough when an auditory-first stimulus condition is presented and the right feeder trough when a visual-first stimulus condition is presented. (B) Schematic of the front wall of the behavioral chamber used for both tasks. The front wall of the chamber consists of a left and right feeder trough and a center nose poke, all outfitted with infra-red (IR) detectors (represented by the red circles within the feeders and nose poke) used for response detection and trial initiation, respectively. The auditory stimulus was delivered from a speaker located above the center nose poke from above the chamber and the visual stimulus was presented from the LED located immediately above the center nose poke. (C) Representative timeline of a single trial for rats trained on either the audiovisual TOJ or SJ task. (D) The experimental timeline for the second experimental series consisting of two different test sessions completed after sham or noise exposure.

Figure 2

Effect of sound intensity on audiovisual temporal order perception. (A) Behavioral performance was plotted as the proportion of responses the rat perceived as “visual-first” (i.e., right feeder trough) for test days completed at 60 dB, 75 dB and 90 dB sound pressure level (SPL). A right-ward shift in the TOJ curve was observed as the intensity of the auditory stimulus increased. For example, at 0 ms stimulus onset asynchrony (SOA) there was an increase in “visual-first” responses at 60 dB SPL when compared to 75 dB SPL (*p < 0.01), and a significant decrease in “visual-first” responses at 90 dB SPL when compared to 75 dB SPL (**p < 0.001). (B) The point of subjective simultaneity (PSS) and (C) the just noticeable difference (JND) were derived from the TOJ task. For PSS, a significant difference was observed between all sound intensities (**p < 0.001), demonstrating a right-ward shift from “auditory-first” responses to “visual-first” responses as the sound intensity increased. For JND, a significance difference was only observed at the lowest sound intensity (i.e., 60 dB SPL), resulting in an increased window of integration (**p < 0.01, ns = not significant). Results are displayed as mean ± standard error of the mean (SEM), n = 10.

Figure 3

Altered auditory- and visual-first performance during TOJ training sessions following noise exposure. (A) Auditory-first performance and (D) visual-first performance pre- and 3 days post-exposure to a loud noise or sham. Following noise exposure there was a slight decrease in auditory-first performance (*p < 0.05, ns = not significant), as well as a significant decrease in visual-first performance (**p < 0.02, ns = not significant). Solid bars represent pre-exposure performance, and patterned bars represent post-exposure performance. Correlation results for (B) auditory-first performance and (E) visual-first performance as a function of final hearing sensitivity (i.e., click thresholds). Gray circles represent the individual data for each rat post-noise exposure. The solid line represents the linear regression line, and the Pearson correlation results along with the significance levels are displayed in the bottom of the panel. Behavioral performance on (C) auditory-first trials and (F) visual-first trials were monitored for 10 days post-exposure. A decrease in performance on auditory-first trials was observed following noise exposure during the first two training sessions (*p < 0.05). Results are displayed as mean ± SEM, n = 9.

Figure 4

Preserved audiovisual temporal perception following adult-onset hearing loss. (A) Test sessions at 75 dB SPL were completed 2 weeks following exposure to a loud noise (i.e., post-noise) or quiet (i.e., post-sham). (B) An additional test session was completed at 90 dB SPL (i.e., post-noise (90 dB SPL)) and compared to the test session at 75 dB SPL (i.e., post-noise (75 dB SPL)), in order to determine if temporal perception remained sensitive to sound intensity. For all test sessions, performance was plotted as the proportion of trials that the rats perceived as “visual-first” (i.e., responded to the right feeder trough; *p < 0.05, **p < 0.007). (C) The PSS and (D) the JND were derived from each of the test sessions (**p < 0.01, ns = not significant). Results are displayed as mean ± SEM, n = 9.

Figure 5

Effect of sound intensity on audiovisual synchrony perception as measured during an SJ task. (A) Behavioral performance was plotted as the proportion of trials the rat perceived as “synchronous” (i.e., left feeder trough) for tests completed at 60 dB, 75 dB and 90 dB SPL. A significant difference was observed at both 60 dB SPL (52.1 ± 3.3%) and 90 dB SPL (85.0 ± 3.3%) when compared to 75 dB SPL (69.0 ± 1.7%; **p < 0.01), indicating that as sound intensity increased, the rate of perceived synchrony also increased when the SOA was less than 100 ms (*p < 0.05). The (B) 50% threshold and (C) 70% threshold were derived from the SJ task. Consistent with the SJ curves, both thresholds showed a significant increase as the intensity of the auditory stimulus increased (**p < 0.01). Results are displayed as mean ± SEM, n = 10.

Figure 6

Hearing loss impaired performance during SJ training sessions. Performance on (A) synchronous and (D) asynchronous trials was compared pre- and 3 days post- exposure to a loud noise or sham. Following noise exposure, a significant decrease in performance on synchronous trials was observed (**p < 0.02, ns = not significant). No difference was observed on asynchronous trials. Solid bars represent pre-exposure performance and patterned bars represent post-exposure performance. Correlation results for (B) synchrony performance and (E) asynchrony performance were plotted as a function of final hearing sensitivity (i.e., click thresholds). Gray circles represent the individual data for each rat post-noise exposure. The solid line represents the linear regression line, and the Pearson correlation results along with the significance levels are displayed in the bottom of the panel. Behavioral performance on (C) synchronous and (F) asynchronous trials were monitored for 10 days following sham and noise exposure. Performance on synchronous trials returned to typical performance within 5 days, whereas performance on asynchronous trials remained consistently impaired across the majority of the training days (*p < 0.05, **p < 0.004). Results are displayed as mean ± SEM, n = 9.

Figure 7

Impaired audiovisual synchrony perception following adult-onset hearing loss. (A) Experimental test sessions for the SJ task at 75 dB SPL were completed 2 weeks following exposure to a loud noise (i.e., post-noise) or quiet (post-sham). (B) An additional test session was completed at 90 dB SPL (i.e., post-noise (90 dB SPL)) and compared to the test session at 75 dB SPL (i.e., post-noise (75 dB SPL)), in order to determine if synchrony perception remained sensitive to sound intensity (*p < 0.05, **p < 0.01). For all test sessions, performance was plotted as the proportion of trials that the rats perceived as “synchronous” (i.e., responded to the left feeder trough). The (C) 50% threshold and (D) 70% threshold were derived from all SJ test sessions. Two weeks following noise exposure, there was a significant increase in the 50% threshold (**p < 0.017), and a modest increase in the 70% threshold (p = 0.08), indicative of a wider window of perceptual binding. Results are displayed as mean ± SEM, n = 9.