An interpretable tinnitus prediction framework using gap-prepulse inhibition in auditory late response and electroencephalogram

BACKGROUND AND OBJECTIVE

Tinnitus is a neuropathological condition that results in mild buzzing or ringing of the ears without an external sound source. Current tinnitus diagnostic methods often rely on subjective assessment and require intricate medical examinations. This study aimed to propose an interpretable tinnitus diagnostic framework using auditory late response (ALR) and electroencephalogram (EEG), inspired by the gap-prepulse inhibition (GPI) paradigm.

METHODS

We collected spontaneous EEG and ALR data from 44 patients with tinnitus and 47 hearing loss-matched controls using specialized hardware to capture responses to sound stimuli with embedded gaps. In this cohort study of tinnitus and control groups, we examined EEG spectral and ALR features of N-P complexes, comparing the responses to gap durations of 50 and 20 ms alongside no-gap conditions. To this end, we developed an interpretable tinnitus diagnostic model using ALR and EEG metrics, boosting machine learning architecture, and explainable feature attribution approaches.

RESULTS

Our proposed model achieved 90 % accuracy in identifying tinnitus, with an area under the performance curve of 0.89. The explainable artificial intelligence approaches have revealed gap-embedded ALR features such as the GPI ratio of N1-P2 and EEG spectral ratio, which can serve as diagnostic metrics for tinnitus. Our method successfully provides personalized prediction explanations for tinnitus diagnosis using gap-embedded auditory and neurological features.

CONCLUSIONS

Deficits in GPI alongside activity in the EEG alpha-beta ratio offer a promising screening tool for assessing tinnitus risk, aligning with current clinical insights from hearing research.

Catechol-O-methyltransferase (COMT) Val158Met Polymorphism and Prepulse Inhibition of the Change-related Cerebral Response

Change-related potentials elicited by an abrupt sound feature's change are attenuated by a leading weak sound (prepulse inhibition: PPI). We investigated whether the PPI index is associated with the catechol-methyltransferase (COMT) Val158Met polymorphism (rs4680), which is involved in the metabolism of dopamine in the prefrontal cortex. Healthy subjects with normal hearing were recruited (n = 70). A train of 100-Hz clicks 650 ms in duration was used. The test stimulus was an abrupt increase in sound intensity (+10 dB) from the baseline (70 dB) provided at 400 ms after the sound onset. Three consecutive clicks at 30, 40, and 50 ms before the change's onset were greater (+3 or +5 dB) from the baseline as a prepulse. The targeting auditory evoked potential component was Change-N1 peaking approx. 130 ms after the change onset. We calculated the inhibition level as the% inhibition of the Change-N1 amplitude by a prepulse. The %PPI in the Met-carriers was significantly greater than that in the Val/Val-individuals. Our results suggest that dopamine might play a role in the PPI of the change-related response. We propose that this index has the potential to identify an intermediate phenotype in psychiatric disorders such as schizophrenia.

Effects of Magnitude of Leading Stimulus on Prepulse Inhibition of Auditory Evoked Cerebral Responses: An Exploratory Study

An abrupt change in a sound feature (test stimulus) elicits a specific cerebral response, which is attenuated by a weaker sound feature change (prepulse) preceding the test stimulus. As an exploratory study, we investigated whether and how the magnitude of the change of the prepulse affects the degree of prepulse inhibition (PPI). Sound stimuli were 650 ms trains of clicks at 100 Hz. The test stimulus was an abrupt sound pressure increase (by 10 dB) in the click train. Three consecutive clicks, weaker (−5 dB, −10 dB, −30 dB, or gap) than the baseline, at 30, 40, and 50 ms before the test stimulus, were used as prepulses. Magnetic responses to the ten types of stimuli (test stimulus alone, control, four types of tests with prepulses, and four types of prepulses alone) were recorded in 10 healthy subjects. The change-related N1m component, peaking at approximately 130 ms, and its PPI were investigated. The degree of PPI caused by the −5 dB prepulse was significantly weaker than that caused by other prepulses. The degree of PPI caused by further decreases in prepulse magnitude showed a plateau level between the −10 dB and gap prepulses. The results suggest that there is a physiologically significant range of sensory changes for PPI, which plays a role in the change detection for survival.

Reliability of Serological Prestin Levels in Humans and its Relation to Otoacoustic Emissions, a Functional Measure of Outer Hair Cells

OBJECTIVES

Serological biomarkers, common to many areas of medicine, have the potential to inform on the health of the human body and to give early warning of risk of compromised function or illness before symptoms are experienced. Serological measurement of prestin, a motor protein uniquely produced and expressed in outer hair cells, has recently been identified as a potential biomarker to inform on the health of the cochlea. Before any test can be introduced into the clinical toolkit, the reproducibility of the measurement when repeated in the same subject must be considered. The primary objective of this study is to outline the test-retest reliability estimates and normative ranges for serological prestin in healthy young adults with normal hearing. In addition, we examine the relation between serum prestin levels and otoacoustic emissions (OAEs) to compare this OHC-specific protein to the most common measure of OHC function currently used in hearing assessments.

DESIGN

We measured prestin levels serologically from circulating blood in 34 young adults (18 to 24 years old) with clinically normal pure-tone audiometric averages at five different timepoints up to six months apart (average intervals between measurements ranged from <1 week to 7 weeks apart). To guide future studies of clinical populations, we present the standard error of the measurement, reference normative values, and multiple measures of reliability. Additionally, we measured transient evoked OAEs at the same five timepoints and used correlation coefficients to examine the relation between OAEs and prestin levels (pg/mL).

RESULTS

Serum prestin levels demonstrated good to excellent reliability between and across the five different time points, with correlation coefficients and intraclass correlations >0.8. Across sessions, the average serum prestin level was 250.20 pg/mL, with a standard error of measurement of 7.28 pg/mL. Moreover, positive correlations (generally weak to moderate) were found between prestin levels and OAE magnitudes and signal-to-noise ratios.

CONCLUSIONS

Findings characterize serum prestin in healthy young adults with normal hearing and provide initial normative data that may be critical to interpreting results from individuals with sensorineural hearing loss. Our results demonstrate reliability of serum prestin levels in a sample of normal-hearing young adults across five test sessions up to 6 months apart, paving the way for testing larger samples to more accurately estimate test-retest standards for clinical protocols, including those involving serial monitoring. The positive correlations between serum prestin and OAE levels, although weak to moderate, reinforce that the source of serum prestin is likely the outer hair cells in the inner ear, but also that serum prestin and OAEs each may also index aspects of biologic function not common to the other.

Electrical field distribution of Change-N1 and its prepulse inhibition

An abrupt change in a sound feature (Test) in a continuous sound elicits an auditory evoked potential, peaking at approx. 100-180 ms (Change-N1) after the change onset. Change-N1 is attenuated by a preceding weak change stimulus (Prepulse), in the phenomenon known as prepulse inhibition (PPI). In this electroencephalographic study, we compared these two indexes among scalp electrodes. Change-N1 was elicited by an abrupt 10-dB increase in sound pressure in repeats of a 70-dB click sound at 100 Hz and was recorded using 22 electrodes in 31 healthy subjects. The prepulse was a 10-dB decrease in three consecutive clicks at 30, 40, and 50 ms before the Test onset. Four stimuli (Test alone, Test with Prepulse, Prepulse alone, and background alone) were presented randomly through headphones at an even probability. The results demonstrated that: (1) Electrodes at the frontal/central midline were reconfirmed to be suitable to record Change-N1; (2) Change-N1 showed right-hemisphere predominance; (3) There was no difference in the %PPI among regions (prefrontal/frontal/central) and hemispheres (midline/left/right); and (4) the Change-N1 amplitude and its PPI at prefrontal electrodes were positively correlated with those at the frontal electrodes. These results support the use of Change-N1 and its PPI as a tool to evaluate the change detection sensitivity and inhibitory function in individuals. The use of prefrontal electrodes can be an option for a screening test.

The gap prepulse inhibition of the acoustic startle (GPIAS) paradigm to assess auditory temporal processing: Monaural versus binaural presentation

The Gap Prepulse Inhibition of the Acoustic Startle Reflex (GPIAS) is a paradigm used to assess auditory temporal processing in both animals and humans. It consists of the presentation of a silent gap embedded in noise and presented a few milliseconds before a startle sound. The silent gap produces the inhibition of the startle reflex, a phenomenon called gap-prepulse inhibition (GPI). This paradigm is also used to detect tinnitus in animal models. The lack of inhibition by the silent gaps is suggested to be indicative of the presence of tinnitus "filling-in" the gaps. The current research aims at improving the GPIAS technique by comparing the GPI produced by monaural versus binaural silent gaps in 29 normal-hearing subjects. Two gap durations (5 or 50 ms), each embedded in two different frequency backgrounds (centered around 500 or 4 kHz). Both low- and high- frequency narrowband noises had a bandwidth of half an octave. Overall, the startle magnitude was greater for the binaural versus the monaural presentation, which might reflect binaural loudness summation. In addition, the GPI was similar between the monaural and the binaural presentations for the high-frequency background noise. However, the GPI was greater for the low-frequency background noise for the binaural, compared to the monaural, presentation. These findings suggest that monaural GPIAS might be more suited to detect tinnitus compared to the binaural presentation.

Comparisons of Auditory Brainstem Responses Between a Laboratory and Simulated Home Environment

Purpose Miniaturization of digital technologies has created new opportunities for remote health care and neuroscientific fieldwork. The current study assesses comparisons between in-home auditory brainstem response (ABR) recordings and recordings obtained in a traditional lab setting. Method Click-evoked and speech-evoked ABRs were recorded in 12 normal-hearing, young adult participants over three test sessions in (a) a shielded sound booth within a research lab, (b) a simulated home environment, and (c) the research lab once more. The same single-family house was used for all home testing. Results Analyses of ABR latencies, a common clinical metric, showed high repeatability between the home and lab environments across both the click-evoked and speech-evoked ABRs. Like ABR latencies, response consistency and signal-to-noise ratio (SNR) were robust both in the lab and in the home and did not show significant differences between locations, although variability between the home and lab was higher than latencies, with two participants influencing this lower repeatability between locations. Response consistency and SNR also patterned together, with a trend for higher SNRs to pair with more consistent responses in both the home and lab environments. Conclusions Our findings demonstrate the feasibility of obtaining high-quality ABR recordings within a simulated home environment that closely approximate those recorded in a more traditional recording environment. This line of work may open doors to greater accessibility to underserved clinical and research populations.

Effect of age on the gap-prepulse inhibition of the cortical N1-P2 complex in humans as a step towards an objective measure of tinnitus

The gap-prepulse inhibition of the acoustic startle reflex has been widely used as a behavioral method for tinnitus screening in animal studies. The cortical-evoked potential gap-induced inhibition has also been investigated in animals as well as in human subjects. The present study aimed to investigate the effect of age on the cortical N1-P2 complex in the gap-prepulse inhibition paradigm. Fifty-seven subjects, aged 20 to 68 years, without continuous tinnitus, were tested with two effective gap conditions (embedded gap of 50- or 20-ms duration). Retest sessions were performed within one month. A significant gap-induced inhibition of the N1-P2 complex was found in both gap durations. Age differently affected the inhibition, depending on gap duration. With a 50-ms gap, the inhibition decreased significantly with the increase in age. This age-inhibition relationship was not found when using a 20-ms gap. The results were reproducible in the retest session. Our findings suggest that the interaction between age and gap duration should be considered when applying the gap-induced inhibition of the cortical-evoked potential as an objective measure of tinnitus in human subjects. Further studies with tinnitus patients are warranted to identify gap duration that would minimize the effects of age and maximize the difference in the inhibition between those with and without tinnitus.

Test-retest reliability of prepulse inhibition paradigm using auditory evoked potentials

Prepulse inhibition (PPI) is a neurological phenomenon in which a weak initial stimulus reduces the level of responses to a subsequent stronger stimulus. Although acoustic startle reflexes are usually used for PPI examinations, recent studies have observed similar phenomena with event-related cortical potentials. In the present study, test-retest reliability of PPI measured using auditory change-related cortical responses was assessed in 35 healthy adults. Four sound stimuli were randomly presented at an even probability; Standard, Test alone, Prepulse alone, and Test + Prepulse. The Standard stimulus was a train of 25-ms tone pulses at 70 dB for 650 ms, while for Test alone and Prepulse alone, the sound pressure was increased to 80 dB at 350 ms and 73 dB at 300 ms, respectively. Measurements were performed twice with at least 7 days separation, and validity was evaluated using intra-class correlation (ICC) for latency, amplitude, and suppression rate of the P50, N100, and P200 components. The results showed high ICC values for the latency and amplitude of nearly all components, except for response to Prepulse alone (0.3-0.6). Furthermore, ICC for suppression rate was greater than 0.5 for the peak-to-peak amplitude. Good reproducibility for N100 and P200 components was obtained with this method. The present results support the PPI paradigm as a reliable tool for clinical measurements of inhibitory functions.

Weaker prepulse exerts stronger suppression of a change-detecting neural circuit

Change-N1 peaking 90-180 ms after changes in a sound feature of a continuous sound is clearly attenuated by a preceding change stimulus (called a "prepulse"). Here, we investigated the effects of a preceding decrease in sound pressure on the degree of inhibition of the subsequent Change-N1 amplitude. Using 100-Hz click train sounds, we obtained Change-N1s from 11 healthy volunteers. The two types of test stimuli were an abrupt 10-dB increase from the baseline (70 dB) and the insertion of a 0.45-ms inter-aural time difference in the middle of the sound. Three consecutive clicks at 30, 40, and 50 ms before the change onset that was used as a prepulse were weaker than the background by 5 or 10 dB. The Change-N1 elicited by the two test stimuli was attenuated more strongly by the weaker prepulse, which was not congruent with the theory that the inhibition of the subsequent sensory/sensory-motor processing depends on the sound pressure level of a prepulse. These results suggest that a change in any type of sound feature elicits a change-related response that is inhibited by any type of preceding change stimulus, which reflects auto-inhibition of the change-responding circuit.

Towards a Pragmatic Approach to a Psychophysiological Unit of Analysis for Mental and Brain Disorders: An EEG-Copeia for Neurofeedback

This article proposes what we call an "EEG-Copeia" for neurofeedback, like the "Pharmacopeia" for psychopharmacology. This paper proposes to define an "EEG-Copeia" as an organized list of scientifically validated EEG markers, characterized by a specific association with an identified cognitive process, that define a psychophysiological unit of analysis useful for mental or brain disorder evaluation and treatment. A characteristic of EEG neurofeedback for mental and brain disorders is that it targets a EEG markers related to a supposed cognitive process, whereas conventional treatments target clinical manifestations. This could explain why EEG neurofeedback studies encounter difficulty in achieving reproducibility and validation. The present paper suggests that a first step to optimize EEG neurofeedback protocols and future research is to target a valid EEG marker. The specificity of the cognitive skills trained and learned during real time feedback of the EEG marker could be enhanced and both the reliability of neurofeedback training and the therapeutic impact optimized. However, several of the most well-known EEG markers have seldom been applied for neurofeedback. Moreover, we lack a reliable and valid EEG targets library for further RCT to evaluate the efficacy of neurofeedback in mental and brain disorders. With the present manuscript, our aim is to foster dialogues between cognitive neuroscience and EEG neurofeedback according to a psychophysiological perspective. The primary objective of this review was to identify the most robust EEG target. EEG markers linked with one or several clearly identified cognitive-related processes will be identified. The secondary objective was to organize these EEG markers and related cognitive process in a psychophysiological unit of analysis matrix inspired by the Research Domain Criteria (RDoC) project.

A Gap Prepulse with a Principal Stimulus Yields a Combined Auditory Late Response

BACKGROUND AND OBJECTIVES

The gap prepulse inhibition of the acoustic startle response has been used to screen tinnitus in an animal model. Here, we examined changes in the auditory late response under various conditions of gap prepulse inhibition.

SUBJECTS AND METHODS

We recruited 19 healthy adults (5 males, 14 females) and their auditory late responses were recorded after various stimuli with or without gap prepulsing. The N1 and P2 responses were selected for analysis. The gap prepulse inhibition was estimated to determine the optimal auditory late response in the gap prepulse paradigm.

RESULTS

We found that the gap per se generated a response that was very similar to the response elicited by sound stimuli. This critically affected the gap associated with the maximal inhibition of the stimulus response. Among the various gap-stimulus intervals (GSIs) between the gap and principal stimulus, the GSI of 150 ms maximally inhibited the response. However, after zero padding was used to minimize artifacts after a P2 response to a gap stimulus, the differences among the GSIs disappeared.

CONCLUSIONS

Overall, the data suggest that both the prepulse inhibition and the gap per se should be considered when using the gap prepulse paradigm to assess tinnitus in humans.

Efficacy of stimulus intensity increases and decreases as inhibitors of the acoustic startle response

The human startle eyeblink response can be inhibited by a change in the stimulus environment briefly before the startling stimulus; both stimulus presentation (prepulse) and cessation of background sound (gap) can result in startle inhibition. More intense prepulses often result in greater inhibition, and this study (N = 53 college students) examined whether graded decreases in sound energy relative to a steady background noise (a "partial gap") would follow this same pattern of inhibition. Embedded in a 65 dB steady background noise were 100 dB white noise startle stimuli preceded at 120 ms on some trials by stimulus intensity increases or decreases of 5, 10, or 15 dB relative to background. Results showed that startle inhibition was graded by amount of change relative to background, such that greater increases or decreases resulted in greater inhibition. Also, increases were more effective startle inhibitors than decreases at equivalent levels of change from background. These results demonstrate that the neural centers responsible for startle inhibition are responsive to both increases and decreases in stimulus intensity, and are sensitive to amount of change, not simply whether a change occurs. These findings may have implications for the development of a screening method for a hearing disorder called tinnitus.

Gap-induced reductions of evoked potentials in the auditory cortex: A possible objective marker for the presence of tinnitus in animals

•

Gap-suppression of startle responses is regularly used as a measure for tinnitus.

•

We studied this phenomenon in auditory cortical evoked potentials in awake animals.

•

Gap-suppression of evoked potentials was also examined following noise exposure.

•

120 dB SPL noise exposure, but not 105 dB, resulted in deficits in gap-suppression.

•

Results are discussed in the context of a potential correlate of tinnitus.

Animal models of tinnitus are essential for determining the underlying mechanisms and testing pharmacotherapies. However, there is doubt over the validity of current behavioural methods for detecting tinnitus. Here, we applied a stimulus paradigm widely used in a behavioural test (gap-induced inhibition of the acoustic startle reflex GPIAS) whilst recording from the auditory cortex, and showed neural response changes that mirror those found in the behavioural tests. We implanted guinea pigs (GPs) with electrocorticographic (ECoG) arrays and recorded baseline auditory cortical responses to a startling stimulus. When a gap was inserted in otherwise continuous background noise prior to the startling stimulus, there was a clear reduction in the subsequent evoked response (termed gap-induced reductions in evoked potentials; GIREP), suggestive of a neural analogue of the GPIAS test. We then unilaterally exposed guinea pigs to narrowband noise (left ear; 8–10 kHz; 1 h) at one of two different sound levels – either 105 dB SPL or 120 dB SPL – and recorded the same responses seven-to-ten weeks following the noise exposure. Significant deficits in GIREP were observed for all areas of the auditory cortex (AC) in the 120 dB-exposed GPs, but not in the 105 dB-exposed GPs. These deficits could not simply be accounted for by changes in response amplitudes. Furthermore, in the contralateral (right) caudal AC we observed a significant increase in evoked potential amplitudes across narrowband background frequencies in both 105 dB and 120 dB-exposed GPs. Taken in the context of the large body of literature that has used the behavioural test as a demonstration of the presence of tinnitus, these results are suggestive of objective neural correlates of the presence of noise-induced tinnitus and hyperacusis.

Behavioral methods for the functional assessment of hair cells in zebrafish

Zebrafish is an emerging animal model for studies on auditory system. This model presents high comparability with humans, good accessibility to the hearing organ, and high throughput capacity. To better utilize this animal model, methodologies need to be used to quantify the hearing function of the zebrafish. Zebrafish displays a series of innate and robust behavior related to its auditory function. Here, we reviewed the advantage of using zebrafish in auditory research and then introduced three behavioral tests, as follows: the startle response, the vestibular-ocular reflex, and rheotaxis. These tests are discussed in terms of their physiological characteristics, up-to-date technical development, and apparatus description. Test limitation and areas to improve are also introduced. Finally, we revealed the feasibility of these applications in zebrafish behavioral assessment and their potential in the high-throughput screening on hearing-related genes and drugs.

The gap-startle paradigm to assess auditory temporal processing: Bridging animal and human research

The gap-prepulse inhibition of the acoustic startle (GPIAS) paradigm is the primary test used in animal research to identify gap detection thresholds and impairment. When a silent gap is presented shortly before a loud startling stimulus, the startle reflex is inhibited and the extent of inhibition is assumed to reflect detection. Here, we applied the same paradigm in humans. One hundred and fifty-seven normal-hearing participants were tested using one of five gap durations (5, 25, 50, 100, 200 ms) in one of the following two paradigms-gap-embedded in or gap-following-the continuous background noise. The duration-inhibition relationship was observable for both conditions but followed different patterns. In the gap-embedded paradigm, GPIAS increased significantly with gap duration up to 50 ms and then more slowly up to 200 ms (trend only). In contrast, in the gap-following paradigm, significant inhibition-different from 0--was observable only at gap durations from 50 to 200 ms. The finding that different patterns are found depending on gap position within the background noise is compatible with distinct mechanisms underlying each of the two paradigms.