Fig 1.
Assumption behind the connection between tinnitus, cochlear synaptopathy and high-frequency hearing loss.
The assumption is that a noise trauma triggers a temporal threshold shift, accompanied by cochlear synaptopathy. With time, the thereby reduced input into the brain stem leads to tinnitus and to spiral ganglion neuron (SGN) degeneration. Progressive effects of synaptopathy and SGN degeneration lead to high-frequency hearing loss under the assumption of higher vulnerability of the basal part of the cochlea. The same timeline might apply to age-related cochlear synaptopathy and shortened by noise overexposure.
Table 1.
Parameters for the Bayesian procedure applied to high-frequency audiometry (HFA), psychophysical tuning curve at 1 kHz (PTC1k), psychophysical tuning curve at the tinnitus frequency (PTCtf), and the tinnitus tuning curve (TTC).
fT indicates the tinnitus frequency.
Table 2.
Summary of the tinnitus frequencies for each participant (IDs) assessed with the tinnitus likeness.
In the PTCtf the control particpants (IDc) have been tested with the frequency of the participant matched.
Fig 2.
Panels A and B: ACALOS results for two exemplary participants.
In panel A, the loudness growth function shows a fast growth for stimulus levels above around 80 dB SPL. The loudness growth functions for all frequencies cross the 25 CU line (most comfortable level) at similar stimulus levels. For the listener in panel B, there is a higher difference between loudness growth across frequency and, compared to the listener in panel A, a more linear growth across the tested input levels. Panel C shows the comparison of the data for the control group (left) and the tinnitus group (right). The difference between the two groups is not significant.
Fig 3.
Exemplary results of MEMR in terms of delta absorbance in different frequency bands for listeners s01 (A) and s02 (B) in the tinnitus group.
Each line represents a different level of stimulation from 75 to 105 dB SPL. The postfix to the listener label indicates the ear tested.
Fig 4.
Two exemplary results of HFA for listeners s01 and s02 (tinnitus group).
The insets show the results of the standard audiometry up to 8 kHz. The central lighter line represents the PTC. The two darker lines represent the limits of the one standard deviation shift. The circles represent the trials in which the target has been heard, and crosses represent the trials in which the target has not been heard.
Fig 5.
Panel A: The psychophysical tuning curve (PTC) at 1 kHz for listener s16 (tinnitus group). Panel B: For the same participant the PTC at the tinnitus frequency (tf), in this case 6 kHz. Panel C: For the same participant the tinnitus tuning curve (TTC).
Fig 6.
Clustering output when including all the 19 predictors and all the participants.
The circles represent control participants, while the squares the tinnitus ones. Different colors represent different clusters.
Fig 7.
Same as Fig 6, but for clustering into three clusters and excluding tinnitus specific predictors.
Table 3.
Resulting parameter values of the clustering analysis with five (upper rows) and three (lower rows) clusters.
For each cluster (Cl, first column), the following parameters are shown: HFA mean (dB HL), MEMR mean, ACALOS mean, age mean, RL500 mean (in ms), PTCtf, PTCtf summary, THI mean, TonLik mean, T/N summary, Mono/Bi (M/B) summary, and TTC summary. In the 5-cluster analysis, all measures were included, while in the 3-cluster analysis the tinnitus measures were excluded.
Fig 8.
The importance analysis for the clustering results in 6.
Fig 9.
Fig 10.
Tinnitus Tuning Curve (TTC) for three exemplary participants: s01 (panel A). s03 (panel B), s09 (panel C). The y-axis shows the level of the masker. The x-axis shows the frequency tested. s01 reported that the tone most similar to his tinnitus had a frequency of 14 kHz, s03 a frequency of 8 kHz and s09 a frequency of 500 Hz.