Skip to main content
NCBI home page
Cerebral Cortex (New York, NY) logoLink to Cerebral Cortex (New York, NY)
. 2024 Jun 17;34(6):bhae220. doi: 10.1093/cercor/bhae220

Structural connectivity changes in unilateral hearing loss

Pascale Tsai 1,2, Timur H Latypov 3,4, Peter Shih-Ping Hung 5,6, Aisha Halawani 7,8,9, Patcharaporn Srisaikaew 10, Matthew R Walker 11, Ashley B Zhang 12, Wanzhang Wang 13, Fatemeh Hassannia 14,15, Rana Barake 16,17, Karen A Gordon 18,19,20, George M Ibrahim 21,22,23,24, John Rutka 25,26, Mojgan Hodaie 27,28,29,30,✉
PMCID: PMC12558133  PMID: 38896551

Abstract

Network connectivity, as mapped by the whole brain connectome, plays a crucial role in regulating auditory function. Auditory deprivation such as unilateral hearing loss might alter structural network connectivity; however, these potential alterations are poorly understood. Thirty-seven acoustic neuroma patients with unilateral hearing loss (19 left-sided and 18 right-sided) and 19 healthy controls underwent diffusion-weighted and T1-weighted imaging to assess edge strength, node strength, and global efficiency of the structural connectome. Edge strength was estimated by pair-wise normalized streamline density from tractography and connectomics. Node strength and global efficiency were calculated through graph theory analysis of the connectome. Pure-tone audiometry and word recognition scores were used to correlate the degree and duration of unilateral hearing loss with node strength and global efficiency. We demonstrate significantly stronger edge strength and node strength through the visual network, weaker edge strength and node strength in the somatomotor network, and stronger global efficiency in the unilateral hearing loss patients. No discernible correlations were observed between the degree and duration of unilateral hearing loss and the measures of node strength or global efficiency. These findings contribute to our understanding of the role of structural connectivity in hearing by facilitating visual network upregulation and somatomotor network downregulation after unilateral hearing loss.

Keywords: auditory, connectome, graph theory, imaging, network

Introduction

Brain connectivity studies have contributed substantially to our understanding of the auditory network, elucidating the underpinnings of the network’s role in auditory processing. In a normal hearing individual, sound inputs from both ears are transmitted bilaterally across the primary and secondary sensory and motor cortices, the temporal cortex, and the prefrontal cortex (Hackett 2011; Lazard et al. 2012; Mišić et al. 2018; Kuiper et al. 2020). These areas are bridged by temporo-premotor and temporo-prefrontal white matter connections, each contributing to unique auditory functions such as speech perception, auditory comprehension, and parallel language processing (Horwitz and Braun 2004; Saur et al. 2010). The auditory network also interacts with other functional networks for multilevel processing of sound and its associated sensory cues (Beer et al. 2013). Functional connectivity studies have explored these cross-modal interactions, aiming to describe the patterns and strength of temporal associations across the whole brain during hearing, or in auditory deprivation states such as unilateral hearing loss (UHL; Litwińczuk et al. 2022). However, structural connectivity in UHL remains underexplored.

Consistently, functional connectivity in UHL has shown to be altered across the whole brain, with significantly stronger connectivity between the auditory network and the visual network, and associated changes through the default mode network across the frontoparietal area (Chen et al. 2020; Shang et al. 2020; Zhang et al. 2016; Zhang et al. 2018b). Particularly, an early study identified superior visual abilities in congenitally deaf cats (Lomber et al. 2010), suggesting a compensatory mechanism for auditory loss through upregulation of the visual network. Weaker connectivity between the auditory and somatomotor networks has also been shown in adults with pre-lingual deafness (Bonna et al. 2021; Andin and Holmer 2022), indicating a downregulation of language–speech coordination normally used to perceive and produce speech (Finkl et al. 2020). Such changes may also occur even with post-lingual UHL; stronger connectivity through the visual network and in the default mode network was shown in adults with UHL (Zhang et al. 2018a), implying cognitive resource reallocation toward effortful listening via strengthened default mode network connectivity (Rönnberg et al. 2011; Yang et al. 2014; Cardin 2016; Choi et al. 2021). Given the robustness of these findings and the inextricable link between functional and structural connectivity, it is highly likely that UHL also manifests as whole brain changes to structural networks and their cross-modal interactions.

Early structural studies indicate a potential link between diffusion tensor imaging metrics and auditory capacity, particularly in cortico–cortical connections, following auditory decline (Qi et al. 2019; Armstrong et al. 2020). Bilaterally poorer hearing in adults was associated with increased mean diffusivity (more free diffusion), reduced fractional anisotropy (poorer white matter integrity), and reduced axial diffusivity, or axon damage, through large-scale white matter tracts that link the frontal, occipital, and temporal lobes (Qi et al. 2019; Armstrong et al. 2020). These findings give strength to the hypothesis that UHL may also significantly alter structural network connectivity across the whole brain. Recent advances in structural imaging, particularly constrained spherical deconvolution tractography (Jeurissen et al. 2014) can build on these studies and support the investigation of structural connectivity to measure the patterns and strength of white matter connections within and between networks in UHL (Sporns 2014).

While diffusion tensor studies can assess the structural integrity of white matter, higher level tractography and connectomics analyses overcome the potential limitations of diffusivity studies. The diffusion-tensor model assumes only 1 orientation of white matter tracts per voxel of biological tissue from diffusion-weighted imaging (DWI; Dell’Acqua and Tournier 2019). The constrained spherical deconvolution model has been used to estimate multiple orientations per voxel (Dell’Acqua and Tournier 2019). With constrained spherical deconvolution, techniques such as anatomically-constrained tractography (ACT) can derive whole brain white matter tracts with a high level of anatomical accuracy (Smith et al. 2012; Bolsterlee et al. 2019; Horbruegger et al. 2019; Dhollander et al. 2021). Structural connectomics is used for further analysis of white matter tracts as the “edges” between gray matter “nodes” (Lynn and Bassett 2019), and surpasses the diffusion-tensor model in its ability to model relative edge strength. Areas with high edge strength form networks, and areas with low edge strength are poorly connected. Graph theory analysis is often applied to structural connectomics to mathematically define node and edge properties. Local graph metrics quantify the strength and efficiency of individual nodes via their edges. Global graph metrics assess the interconnectedness of all nodes (Lynn and Bassett 2019). These analyses can significantly advance our knowledge of structural connectivity in UHL, both within and between networks and through the whole brain connectome. Further, they represent validated approaches in structural connectomics, utilizing graph theory to investigate connectivity in health and disease (Batalle et al. 2017; Kamagata et al. 2018; Zhang et al. 2011).

The aim of the present study was to investigate to what extent whole brain intra- and inter-network structural connectivity changes, measured by edge strength and graph theory metrics, might relate to UHL. A data-driven, whole brain approach was taken to identify connectivity changes that may, for the first time, point to structural alterations outside of the auditory network that are implicated in UHL. Data were collected from patients with unilateral acoustic neuromas (ANs), as these tumors result in ipsilateral UHL and are diagnosed through MRI scans and audiometry tests. Specifically, study aims were to: (i) estimate pair-wise edge strength through the whole brain connectome to define differences in structural connectivity between UHL patients and healthy control (HCs) using constrained spherical deconvolution and ACT, (ii) characterize structural connectivity using local and global graph theory metrics to explore key nodes and networks, and (iii) correlate graph theory metrics to audiometry data, taking the duration and degree of UHL into account. Study hypotheses were that edge strength and graph theory metrics of the temporal, prefrontal/frontal, premotor/motor, and visual areas would significantly associate with UHL and define key contributory networks, and that UHL duration and degree would associate with graph metrics of structural connectivity.

Materials and methods

Ethics

This retrospective AN study was approved by the University Health Network (UHN) Research Ethics Board. All patient data analyzed in this study was retrospective, requiring no active participation. Thus, no informed consent was required for this study. The UHN Research Ethics Board also approved the collection of HC data and imaging acquisition from these individuals. Written informed consent was obtained from all HCs. All imaging data for all participants included in this study was anonymized prior to analysis.

Participants

Patients with a unilateral AN (either left- [L-AN] or right-sided [R-AN]) and UHL, and HCs age- and sex-matched to each patient group were studied. Recruitment criteria for all HCs included the absence of any neurological conditions or other self-reported health conditions. HCs did not report hearing issues, and therefore were considered to have normal hearing. Exclusion criteria were patients with bilateral ANs due to neurofibromatosis type II, participants with previous surgical treatments (either for the AN or for comorbidities such as cancer, cognitive, or motor disorders), or those who had hydrocephalus or pronounced ventricular enlargement. Most of the AN group had ipsilateral sensorineural hearing loss and tinnitus. Given the high co-occurrence of UHL and tinnitus in adults with ANs (Moffat et al. 1998; Baguley et al. 2001; Foley et al. 2017) and inherent presence of tinnitus in association with this patient group, we did not exclude patients with tinnitus from our study.

Hearing was assessed by pure tone audiometry (PTA) and word recognition score (WRS) tests on the same day before surgery in the AN group. The WRS task presents words to each ear of a participant and the percentage of words repeated back correctly is scored (Schlauch et al. 2014). Thus, WRS data provided percent accuracy of word understanding at the most comfortable listening levels in each ear. UHL duration was calculated as the time between a patient’s magnetic resonance imaging (MRI) scan and their audiometry test date. Of note, patients may have noticed UHL earlier than their recorded audiograms. Using this time point allowed for a metric of internal consistency in the definition of UHL duration. Thus, data from each patient’s audiogram accompanying their MRI was chosen as a consistent time point. Side of hearing was defined as either ipsilateral or contralateral to the AN. Ipsilateral high-frequency average hearing thresholds were calculated as the average PTA hearing threshold (dB) between 4 and 8 kHz for each subject and used for comparison to graph theory metrics. Ipsilateral WRS data was also used for comparison to graph theory metrics.

MRI data acquisition

Each patient and HC underwent the same clinical imaging protocol with 3 Tesla GE Signa HDx MRI scans and an 8-channel head coil. Single-shell DWI data were acquired with the following parameters: 60 directions, spin echo echo-planar imaging (EPI) sequence, 1 B0, b = 1,000 s/mmb2, array spatial sensitivity encoding technique, voxel size = 0.94 × 0.94 × 3 mmc3, matrix = 256 × 256, TR = 12 s, TE = 87.3 ms, flip angle = 90°, field of view (FOV) = 240 mm. Gadolinium contrast-enhanced T1-weighted (T1w) anatomical images with a 3D fast spoiled gradient-echo sequence (repetition time/echo time [TR/TE] = 9.0/3.7 ms, matrix = 512 × 512, voxel size = 0.39 × 0.39 × 1 mmc3, flip angle = 12°, FOV = 200 mm, inversion time [TI] = 450 ms). T2-weighted fast imaging employing steady-state acquisition scans were also obtained using an inversion recovery fast-spoiled gradient recalled sequence for our neuroradiology expert (A.H.) to assess and ensure concordance with the expected radiological signatures for ANs. Software tools used for image processing and connectome analysis are described in Supplementary Material.

DWI and T1w image preprocessing, tractography, and connectome generation

The details of the processing are given in detail in Supplementary material. In brief, DWI data was corrected for eddy-currents and motion distortions. B0 images were extracted from the DWI data, skull-stripped, and up-sampled to T1 voxels. Raw T1w images were parcellated into 70 cortical and 14 subcortical gray matter structures using the Desikan–Killiany atlas (Desikan et al. 2006). T1w images were skull-stripped and registered to the B0 image for each subject. Transformation matrices generated from this registration were used to transform the parcellation images into DWI space. Constrained spherical deconvolution was used to estimate fiber orientation distributions (FODs) per voxel on each subject’s DWI data. ACT was used for whole brain fiber-tracking with the anatomical parcellation image and FOD image per subject. Spherical deconvolution-informed filtering of tractograms 2 (SIFT2) was used for tractogram filtering. Correlation matrices representing the structural connectomes were generated per subject with the connectivity information for 84 cortical and subcortical nodes from the streamline weights generated with ACT and SIFT2.

Definition of edge strength, network normalization, and thresholding

The contribution of inter-nodal streamline weights to edge strength was normalized to the total sum of weights across all network streamlines (10 million) at the whole brain level, given by Equation 1 (Batalle et al. 2017):

graphic file with name DmEquation1.gif (1)

where from nodes i to j, wX is the weighting factor (here as NSD), and eij is the sum of weights contributed by the individual streamlines, that is divided by the sum of weights contributed by all streamlines from node k to node l (ekl) in the whole brain tractogram. An in-house algorithm with MATLAB was used for this computation.

Each subject’s normalized network was thresholded to a proportion of connections (actual connections/potential connections) corresponding to network densities of 0.1, 0.15, 0.2, 0.25, 0.3, 0.35, 0.4, 0.45, and 0.5, following a similar approach to that used in previous studies (Batalle et al. 2017; Kamagata et al. 2018; Zhang et al. 2011). To avoid bias from using a single threshold, graph theory metrics were examined across this threshold range. To identify variability across the range of network densities per subject, each subject’s networks were averaged across this range and correlated to the NSD-weighted networks (Supplementary material). Averaged connectivity matrices per group were noted to be highly visually similar to the NSD-weighted networks (Supplementary material) with high (>0.9) correlation coefficients between networks and P-values close to 0. This suggested that our connectivity findings would be similar from either NSD-weighted or thresholded networks.

Graph theory analysis

One global metric and 1 local metric were computed: global efficiency and node strength, respectively (Table 1). To account for the hypothesized connectivity changes across multiple networks through the whole brain in UHL, global efficiency (efficiency_wei.m) was calculated as a metric of parallel information flow between either disconnected or non-sparse networks or both (Zhao et al. 2015). Global efficiency is inversely proportional to the characteristic path length, where the shorter the average shortest inter-nodal path length, the stronger the global efficiency. Node strength (strengths_und.m) was calculated as the sum of the weights of edges connected to a node, representing the network hub property or importance of each node in a network (Sethi et al. 2017). It was chosen to identify the relative contributions of individual brain regions to network connectivity, in support of our hypothesis that the temporal, prefrontal/frontal, premotor/motor, and visual areas would significantly associate with UHL. For comparison of variability across densities, global efficiency values for each network density were extracted and compared between groups (L-AN, R-AN, HC) (Supplementary material). Highly similar group differences in global efficiency were observed across network densities, confirming the similarity in connectivity findings across thresholds. With the similarity between NSD-weighted and thresholded matrices noted above, we restricted our remaining analyses to the NSD-weighted networks.

Table 1.

Graph theory metrics. We assessed 1 local metric and 1 global metric.

Metric Definition
Global (network)
Global efficiency:
Inline graphic, where the global efficiency Inline graphic for a weighted undirected connectome is the average of the inverse shortest path length, Inline graphic, N is the set of all nodes in the connectome (i, j ∈ N), and n is the number of nodes in connectome N. 		The average of the inverse shortest path length in the connectome, where the shortest path length is the average minimum number of connections that link any two nodes. Represents the capability for parallel information propagation in the whole brain connectome.
Local (nodal)
Node strength:
Inline graphic, where Inline graphic is a 1D vector with n elements and n is the number of nodes in the network N. Inline graphic is the strength of node i, equal to the sum of weights (Inline graphic) of all the connections of node i. 		The sum of the weights of links connected to a node. Represents how connected 1 node is to others based on the strengths of each of its connections.
Open in a new tab

Statistical analysis

Python was used to compare edge strength, defined by pair-wise NSDs across each subject’s connectome between groups: (i) L-AN compared to HCs, (ii) R-AN compared to HCs, and (iii) L-AN compared to R-AN. Welch’s t-tests were used to compare the 84 pair-wise NSDs per subject between groups for each of the 3 tests, and the false-discovery rate (FDR) approach to correct the corresponding p-values (significance was defined as P < 0.05) (Benjamini and Hochberg 1995).

Python was used to convert the connectivity matrices generated on MATLAB into formats readable in R software. All following statistical analyses were performed using R v 3.6.3 software (R Core Team, 2013). Linear regression models with age and sex as covariates were used to compare graph theory metrics between groups. The general format (Equation 2) for these models was as follows:

graphic file with name DmEquation2.gif (2)

where Y was the dependent variable indicating the graph theory metric (global efficiency or node strength), and the independent variables included tumor side (left for L-AN, right for R-AN, or “no tumor” for the HCs), age, and sex. In comparing audiometry variables to graph theory metrics, this general model was modified to include either ipsilateral hearing score or the duration of hearing loss as another fixed covariate. Least-squares means were used for post-hoc comparisons of factors in the linear model used to compare global efficiency (the network-level metric) between groups, with the Satterthwaite method for correcting degrees of freedom. When comparing the edge-wise graph theory metric (node strength) between groups, the general linear model was run separately for each node. The corresponding edge-wise statistics were corrected for multiple comparisons using the FDR approach. The FDR approach was also used to correct for multiple comparisons between linear mixed models per sound frequency comparing the PTA thresholds between ipsilateral and contralateral ears.

Results

Participant demographics

This study included a total of 56 subjects: 37 UHL patients at the Toronto Western Hospital (L-AN: 19, 10F, 9 M, mean age ± SD: 54.1 ± 6.1 yr; R-AN: 18, 11F, 7 M, 52.7 ± 9.5) and 19 HCs (10F, 9M, 52.2 ± 14.2) tested at the Toronto Western Hospital (Fig. 1). UHL patients were diagnosed with vestibular schwannomas and underwent radiosurgery for treatment of their tumors.

Fig. 1.

Fig. 1

Open in a new tab

Age distribution by group. The distribution of age did not significantly differ between groups F(2, 53) = 0.16.

Hearing assessment

PTA thresholds at frequencies of 0.25, 0.5, 1, 2, 4, and 8 kHz were available from each ear in 32/37 adults. WRS data was available in 27/37 adults. A total of 26/32 patients had ipsilateral high-frequency average (between 4 and 8 kHz) PTA thresholds above 25 dB (mild to profound hearing loss, 27.5 to 95 dB). Only 7/32 had contralateral high-frequency average PTA thresholds at or above 25 dB (25 to 62.5 dB). This suggested that a minority of patients had poor hearing bilaterally. Ipsilateral high-frequency average PTA thresholds were significantly higher than contralateral thresholds for all patients [F(1, 30) = 15.4, P = 0.00048]. These data are presented per PTA frequency in Fig. 2. There were no significant differences in ipsilateral [F(1, 30) = 0.001] or contralateral [F(1, 30) = 0.009] high-frequency average PTA thresholds between L-AN and R-AN. Ipsilateral WRS data were similar to contralateral WRS data [F(1, 25) = 0.12, P = 0.74], and there were no differences in ipsilateral [F(1, 25) = 1.42] or contralateral [F(1, 25) = 1.94] WRS data between L-AN and R-AN (Fig. 2). A full summary of clinical data is provided in Table 2.

Fig. 2.

Fig. 2

Open in a new tab

Average A) PTA hearing thresholds and B) WRS data. A) Circles represent ipsilateral average PTA thresholds and triangles represent contralateral average PTA thresholds for L-AN and R-AN. Hearing was significantly better contralaterally across all PTA frequencies [0.25 kHz: T(62) = −3.3, P = 0.001; 0.5 kHz: T(62) = −3.5, P = 0.0008; 1 kHz: T(62) = −5.1, P = 3.96e-06; 2 kHz: T(62) = −5.7, P = 4.31e-07; 4 kHz: T(62) = −6.1, P = 6.34e-08; 8 kHz: T(62) = −4.9, P = 8.47e-06], including after FDR-correction (pFDR <0.05). B) Contralateral and ipsilateral WRS data did not differ across patients [F(1, 25) = 0.12, P = 0.74].

Table 2.

Participant demographics. Numbers represent mean ± SD values for continuous variables, and group counts for categorical variables. †Handedness information is presented for each group out of the subjects whose data we could obtain, including for all 18/18 R-AN, 17/19 L-AN, and 9/19 of the HCs.

Variable R-AN (n = 18) L-AN (n = 19) HC (n = 19) Statistic
*P < 0.05
Sex: female/male (% male) 11/7 (38.9%) 10/9 (47.4%) 10/9 (47.4%) χb2(2) = 0.36
Handedness (L/R/cross-dominant)† 5/13/0 2/15/0 0/8/1 χb2(4) = 7.44
Ipsilateral sensorineural hearing loss (Y/N) 16/2 17/2 — χb2(1) = 1.78e-30
Tinnitus (Y/N) 14/4 14/5 — χb2(1) = 1.13e-31
Duration of hearing loss (years) 1.35 ± 1.38 0.94 ± 0.64 — F(1, 30) = 1.20
Open in a new tab

Edge strength differences between patients and controls

Comparison of edge strength between groups (Welch’s t-tests) showed significant differences between patients and controls in the occipital (visual), medial frontoparietal (default), pericentral (somatomotor), lateral frontoparietal (control), and cerebellar networks FDR-corrected P < 0.05; Fig. 3).

Fig. 3.

Fig. 3

Open in a new tab

Group comparisons of edge strength. Lines indicate where edge strength was either increased [Welch’s t-tests, FDR-corrected P-values (pFDR) <0.05] or reduced between A) L-AN or B) R-AN and controls. Edge strength was increased in patients relative to controls between the cerebellar and visual networks, and reduced between the default, somatomotor and control, and default and somatomotor networks. The full list of region abbreviations is provided in the Supplementary Material.

The L-AN group demonstrated significant left-lateralized increases in edge strength, particularly in nodes connecting the cerebellar and visual networks. Conversely, the R-AN group exhibited significantly increased edge strength bilaterally within the visual network.

Edge strength was significantly reduced in the L-AN group in nodes through the default network bilaterally. Edge strength was also significantly reduced in left and right node pairs between the default and somatomotor networks, and right nodes between the somatomotor and control networks. Reduced edge strength in the R-AN group was right-lateralized to nodes that paired the default and somatomotor networks. Full details of this analysis, including node to network associations are provided in Table 3.

Table 3.

Regions between which edge strength was a) significantly increased or b) significantly reduced in patients compared to controls. After FDR-correction, Welch’s t-tests identified 6 networks through which edge strength significantly differed between patients and controls. Positive signs of t-values indicate higher edge strength in patients, negative signs reflect the opposite. ITG = inferior temporal gyrus, L = left, R = right.

Connection Comparison t value pFDR value Networks
AN > HC
L-CER, L-LG L-AN vs HC 5.17 0.045 Cerebellar and visual
L-CER, L-FG 5.30 0.045 Cerebellar and visual
R-LG, L-LG R-AN vs HC 6.76 0.005 Visual
AN < HC
L-SFG, R-POP L-AN vs HC −5.23 0.045 Default
L-SFG, L-PoCG −4.52 0.045 Default and somatomotor
R-PoCG, R-ITG −4.77 0.045 Somatomotor and control
R-MTG, R-PoCG −5.02 0.045 Default and somatomotor
R-MTG, R-PoCG R-AN vs HC −5.43 0.030 Default and somatomotor
Open in a new tab

Graph metrics and associations with audiometry

Increased overall global efficiency was observed in patients compared to controls [F(2, 51) = 7.95, P = 0.00099; Fig. 4]. Post-hoc testing revealed significantly increased global efficiency in L-AN [t(51) = 3.80, P = 0.001] and R-AN [t(51) = 2.74, P = 0.023] relative to controls. No differences were observed between L-AN and R-AN [t(51) = 1.01, P = 0.573].

Fig. 4.

Fig. 4

Open in a new tab

Global efficiency values A) per group, compared to B) ipsilateral high-frequency average PTA thresholds, C) ipsilateral WRS data, and D) UHL duration. A) Asterisks (* or **) indicate significant (P < 0.05 and P < 0.01, respectively, after post-hoc comparisons with least-squares means of the linear regression model) group differences in mean global efficiency. Global efficiency values were significantly increased in both L-AN and R-AN relative to controls. B) There were no relationships between global efficiency and ipsilateral PTA hearing thresholds, C) ipsilateral WRS scores, or D) UHL duration. Lines represent linear regression models, and shaded bands indicate 95% confidence intervals of each line.

Linear regression analysis identified no significant interaction between ipsilateral high-frequency average PTA hearing thresholds and global efficiency [F(1, 27) = 1.85, P = 0.19], between ipsilateral WRS scores and global efficiency [F(1, 22) = 0.85, P = 0.37], or between UHL duration and global efficiency [F(1, 27) = 0.16, P = 0.69; Fig. 4].

Linear regression models were used to compare node strength between groups (L-AN, R-AN, and HCs) at all connectome nodes (Fig. 5). Node strength was significantly increased in the L-AN group compared to controls bilaterally through the default and visual networks (pFDR < 0.05, details given in Table 4). It was significantly reduced in this group bilaterally through nodes of the somatomotor network. Node strength was significantly increased in the R-AN group through the visual network bilaterally, and in a left node of the default network (Table 4). It was significantly reduced in a left node of the somatomotor network. These results are fully summarized in Table 4, including network to node associations. There were no significant differences in node strength between L-AN and R-AN (Fig. 5). PFDR-values from the linear models and mean node strength values at nodes for which there were significant group differences are plotted in Fig. 5.

Fig. 5.

Fig. 5

Open in a new tab

Group comparisons of node strength. A) Glass brains show where significant differences in node strength were identified between patients and controls. B and C) Line heights represent pFDR-values from linear regression models comparing node strength between the 3 groups per node for B) all 42 left-sided nodes, and C) all 42 right-sided nodes. Dots within the rectangles below the horizontal dashed lines at x = 0.05 signify where node strengths significantly differed (pFDR <0.05) between groups. D and E) Mean node strength values per group across each node in which significant group differences were identified. Asterisks indicate significant (pFDR <0.05) group differences in node strength. *P < 0.05; **P < 0.01; ***P < 0.001.

Table 4.

Group comparisons of node strength through the left and right hemispheres. T-values for which there were significant group differences (L-AN compared to HCs and R-AN compared to HCs) in node strength are provided. All listed t-values had a significance of pFDR < 0.05. Positive signs of t-values indicate that node strength is higher in the AN group relative to the HC group, negative signs reflect the opposite. NS = not significant, FP = frontal pole.

Node L-AN vs HC (t) L-AN vs HC (P) R-AN vs HC (t) R-AN vs HC (P) Network
Right hemisphere
PoCG −3.42 0.046 NS NS Pericentral (somatomotor)
ICG 3.37 0.046 NS NS Medial frontoparietal (default)
LG 4.61 0.003 4.16 0.008 Occipital (visual)
Left hemisphere
TP 3.24 0.030 4.53 0.002 Medial frontoparietal (default)
PrCG −4.47 0.002 −3.24 0.030 Pericentral (somatomotor)
PoCG −4.05 0.005 NS NS Pericentral (somatomotor)
LG 5.26 0.0004 3.32 0.030 Occipital (visual)
PCAL 3.54 0.018 3.55 0.018 Occipital (visual)
Open in a new tab

Relationships between node strength and audiometry variables with separate linear models at each node were assessed. L-AN and R-AN were grouped and assessed in these models together as there were no differences in node strength between them. After FDR-correction for multiple comparisons, there were no linear relationships between node strength and ipsilateral WRS scores, UHL duration, or ipsilateral PTA thresholds (data not shown).

Discussion

One of the fundamental challenges in understanding the impact of UHL lies in deciphering how whole brain structural connectivity may reorganize. Findings of this study indicate significant changes in structural connectivity within the visual and somatomotor networks in UHL. Increased connectivity within the visual network and stronger connections between the visual and cerebellar networks were observed. In contrast, the somatomotor network displayed reduced edge and node strength, alongside reduced connectivity to the default and control networks. Furthermore, an elevation in global efficiency suggests an overall enhancement in the brain’s capacity for information integration in UHL. Notably, audiometric measures of UHL duration and degree showed no correlation with either node strength or global efficiency. This study represents, to our knowledge, a novel application of structural connectomics and graph theory analysis in UHL. It provides insights into potential adaptive mechanisms, such as the visual network’s strengthened connectivity and hypothesized resource reallocation from the somatomotor network, typically involved in speech processing, to the visual network for compensatory purposes. The present findings contribute significantly to the existing knowledge, highlighting the pivotal roles of the visual and somatomotor networks in the brain’s reorganization due to UHL.

Visual and somatomotor networks as key contributors in UHL

In this study, the importance of individual nodes in contributing to structural network strength, and the strengths of their connections to other nodes either within or across networks were simultaneously examined. This was done by using edge strength to examine pair-wise NSD between nodes across the whole brain connectome in UHL, and node strength to measure local interconnectedness of nodes. This is a novel approach as prior work examined either only functional connectivity in UHL, functional connectivity in bilateral hearing loss, or provided a limited characterization of connectivity (Zhang et al. 2018a; Wei et al. 2021; Zou et al. 2021; Ponticorvo et al. 2022). For instance, although previous studies have demonstrated that visual network functional connectivity is upregulated in states of auditory deprivation, the present approach provides examination of structural changes both within the visual network and across its connections. Additionally, the whole brain analysis of global efficiency, local node strength, and edge strength in this study allowed for examination of non-sparse and disconnected networks that may contribute to UHL, rather than focusing on the dominant networks alone.

The observations from this study highlight that edge strength was significantly increased between the bilateral lingual gyri of the visual network, and from the left fusiform and lingual gyri of the visual network to the left cerebellum (CER), or cerebellar network. Secondly, node strength, the graph metric of local interconnectedness, was increased at the lingual gyrus (LG) bilaterally and at the left pericalcarine cortex (PCAL)—primary visual cortex—of the visual network. These findings were irrespective to the side of UHL as there were no significant differences in edge strength or node strength between L-AN and R-AN. These data are in accordance with previous functional imaging studies (Zhang et al. 2018a; Wei et al. 2021; Zou et al. 2021; Ponticorvo et al. 2022) and with the study hypothesis that visual areas would be key contributors to structural network connectivity changes in UHL.

The visual network, including the fusiform gyrus (FG), LG, and PCAL is involved in multisensory perception, including spatial awareness, object recognition, and goal-oriented movement (Kreutzer et al. 2010; Palejwala et al. 2020; Ponticorvo et al. 2022). Combined with the studies above (Zhang et al. 2018a; Wei et al. 2021; Zou et al. 2021; Ponticorvo et al. 2022), we postulate that augmentation of visual performance via strengthening of white matter connections through the visual network and to the cerebellar network serves to compensate for UHL. The CER particularly is involved in motor coordination, including voluntary eye movement, and in visual attention (Robinson and Fuchs 2001; Brissenden and Somers 2019). In normal hearing, dorsomedial cerebellar to frontoparietal functional connections service visual attention and working memory (Brissenden and Somers 2019). The strengthening of visual to cerebellar network connections in the present study uniquely suggests that plasticity regulates increased visual attentional processing in an auditory deprived state.

A reduction in connectivity within the somatomotor network in UHL, with similar patterns in both L-AN and R-AN, was observed. This was characterized by reduced edge strength between the bilateral postcentral gyrus (PoCG) of the somatomotor network and various nodes of the control and default mode networks, including the left superior frontal gyrus (SFG), right isthmus of the cingulate gyrus (ICG), and right middle temporal gyrus (MTG). Furthermore, a reduction in node strength was noted in the bilateral PoCG and left precentral gyrus (PrCG). These results suggest a downregulation of multiple pericentral–medial and frontoparietal–lateral networks in UHL. Our findings also propose that speech perception and production, primarily facilitated by the somatomotor network in normal hearing, may be compromised in UHL, likely due to an increased reliance on visual information. This is supported by the loss of effective connectivity between the somatomotor and auditory networks in UHL, potentially leading to a compensatory upregulation of the visual network.

Previous functional imaging studies have indicated reduced connections between the somatomotor, default mode, and control networks in UHL, implying reduced effective information flow and processing (Falkenberg et al. 2011; Schmidt et al. 2013; Andin and Holmer 2022). The present results corroborate these findings and contribute to the understanding of how speech perception and production, which typically depend on strong synchronization between the somatomotor and auditory networks (Hickok et al. 2009; Price et al. 2011), are altered in UHL.

Associations between the default mode network and UHL

Significant associations between the default mode network and UHL were identified. There was reduced edge strength between the right pars opercularis (POP) and left SFG, and increased node strength at the temporal pole (TP) and at the ICG. Reduced connectivity between the POP and SFG supports our hypothesis of a reduction in speech and language processing and production in UHL, given the role of the POP in language production, phonological processing, and verbal working memory, and the SFG in working memory. Increased node strength at the TP is in-line with the study hypothesis that visual processing is heightened to compensate for UHL, as this structure is involved in visual processing for complex objects and face recognition (Herlin et al. 2021). Increased node strength at the ICG may suggest an association between UHL, cognitive function, and emotion, as this region may be involved in memory, depression, and pain processing (McLaren et al. 2016; Calati et al. 2018).

Greater global efficiency in UHL

Global efficiency was significantly increased in UHL, with no significant differences between L-AN and R-AN. We theorize that in UHL, the compensatory response to the loss of inputs in 1 part of the brain’s system involves an increase in global efficiency, leading to more targeted connectivity within and across networks, especially the visual and somatomotor networks. This adaptation in global efficiency corresponds with the increased strength and connectivity observed in our study, and is in line with previous functional connectivity research reporting similar increases in global efficiency in cases of age-related bilateral hearing loss (Ponticorvo et al. 2022) and unilateral sudden sensorineural hearing loss (Xu et al. 2016).

Graph metrics with UHL degree and duration

No significant relationships between UHL degree or duration with node strength or with global efficiency were identified. Reasons that significant associations with the audiometry and graph metrics were not observed might include that: (i) UHL duration only manifests in changes to structural connectivity through the brain after a length of time longer than any durations from the participants included in this study, (ii) the sample size of only 37 adults with ANs was too small to denote a relationship between UHL duration and graph metrics, (iii) the subject cohort did not present with a wide enough range of durations of hearing loss to identify a significant relationship between the longitudinal progression of hearing loss and graph metrics, (iv) the relationship between UHL duration and changes to structural connectivity through the brain is non-linear, and was unremarkable with the present methodology, or (v) complete audiometry data were not available for all UHL patients. Additionally, our whole brain analysis of large-scale structural networks may have masked smaller associations limited to regions of the auditory network. A seed-based analysis of structural connectivity in UHL might identify audiometry associations with regions of the auditory network and build upon our work.

Conclusion

The present study reveals the significant impact of UHL on structural connectivity within the whole brain connectome, marking a novel understanding in this field. The observed increase in visual network connectivity and decrease in somatomotor network connectivity implies the crucial role of auditory compensation and the association with communication challenges in UHL. Results also indicate no correlation between UHL duration and degree with structural connectivity metrics. Further research is essential to expand on these findings, including combined structural–functional analysis and a larger sample size.

Supplementary Material

supplementarymaterial_bhae220
supplementarymaterial_bhae220.docx (4.6MB, docx)

Contributor Information

Pascale Tsai, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada; Institute of Medical Science, University of Toronto, 6 Queen's Park Cres, Toronto, Ontario M5S 3H2, Canada.

Timur H Latypov, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada; Institute of Medical Science, University of Toronto, 6 Queen's Park Cres, Toronto, Ontario M5S 3H2, Canada.

Peter Shih-Ping Hung, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada; Institute of Medical Science, University of Toronto, 6 Queen's Park Cres, Toronto, Ontario M5S 3H2, Canada.

Aisha Halawani, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada; Division of Neuroradiology, Joint Department of Medical Imaging, Toronto Western Hospital, University Health Network, 399 Bathurst St, Toronto, Ontario M5T 2S8, Canada; Department of Medical Imaging, Ministry of the National Guard—Health Affairs, C967+PRM, King Abdul Aziz Medical City, Jeddah 22384, Saudi Arabia.

Patcharaporn Srisaikaew, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada.

Matthew R Walker, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada.

Ashley B Zhang, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada.

Wanzhang Wang, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada.

Fatemeh Hassannia, Department of Otolaryngology—Head and Neck Surgery, University of Toronto, 600 University Ave, Toronto, Ontario M5G 1X5, Canada; Temerty Faculty of Medicine, University of Toronto, 1 King's College Cir, Toronto, Ontario M5S 1A8, Canada.

Rana Barake, Department of Otolaryngology—Head and Neck Surgery, University of Toronto, 600 University Ave, Toronto, Ontario M5G 1X5, Canada; Temerty Faculty of Medicine, University of Toronto, 1 King's College Cir, Toronto, Ontario M5S 1A8, Canada.

Karen A Gordon, Institute of Medical Science, University of Toronto, 6 Queen's Park Cres, Toronto, Ontario M5S 3H2, Canada; Department of Otolaryngology—Head and Neck Surgery, University of Toronto, 600 University Ave, Toronto, Ontario M5G 1X5, Canada; Department of Communication Disorders, The Hospital for Sick Children, 555 University Ave, Toronto, Ontario M5G 1X8, Canada.

George M Ibrahim, Institute of Medical Science, University of Toronto, 6 Queen's Park Cres, Toronto, Ontario M5S 3H2, Canada; Temerty Faculty of Medicine, University of Toronto, 1 King's College Cir, Toronto, Ontario M5S 1A8, Canada; Division of Neurosurgery, Department of Surgery, University of Toronto, 149 College St, Toronto, Ontario M5T 1P5, Canada; Institute of Biomedical Engineering, University of Toronto, 164 College St, Toronto, M5S 3G9 Ontario M5S 3G9, Canada.

John Rutka, Department of Otolaryngology—Head and Neck Surgery, University of Toronto, 600 University Ave, Toronto, Ontario M5G 1X5, Canada; Temerty Faculty of Medicine, University of Toronto, 1 King's College Cir, Toronto, Ontario M5S 1A8, Canada.

Mojgan Hodaie, Krembil Research Institute, University Health Network, 60 Leonard Ave, Toronto, Ontario M5T 0S8, Canada; Institute of Medical Science, University of Toronto, 6 Queen's Park Cres, Toronto, Ontario M5S 3H2, Canada; Temerty Faculty of Medicine, University of Toronto, 1 King's College Cir, Toronto, Ontario M5S 1A8, Canada; Division of Neurosurgery, Department of Surgery, University of Toronto, 149 College St, Toronto, Ontario M5T 1P5, Canada.

Author contributions

Pascale Tsai (Conceptualization, Methodology, Software, Formal analysis, Investigation, Data curation, Visualization, Writing – original draft, Writing – review & editing), Timur H. Latypov (Methodology, Software, Formal analysis, Visualization, Writing – review & editing), Peter S. Hung (Methodology, Software, Formal analysis, Visualization, Writing – review & editing), Aisha Halawani (Conceptualization, Validation, Data curation, Writing – review & editing), Patcharaporn Srisaikaew (Visualization, Validation, Writing – review & editing), Matthew R. Walker (Data curation, Writing – review & editing), Ashley B. Zhang (Data curation, Writing – review & editing), Wanzhang Wang (Data curation, Writing – review & editing), Fatemeh Hassannia (Data curation, Writing – review & editing), Rana Barake (Data curation, Writing – review & editing), Karen A. Gordon (Conceptualization, Writing – review & editing), George M. Ibrahim (Conceptualization, Writing – review & editing), John Rutka (Writing – review & editing), Mojgan Hodaie (Conceptualization, Methodology, Supervision, Project administration, Writing – review & editing).

Funding

This work was supported by awards to P.T. from the Natural Science and Engineering Research Council and the Ontario Student Opportunity Trust Fund. P.S.H. acknowledges support from the Canadian Institutes of Health Research (grant number: GSD157876).

Conflict of interest statement: None declared.

Data availability

Anonymized data and code used for this study is available upon reasonable request from the corresponding author.

References

  1. Andin J, Holmer E. Reorganization of large-scale brain networks in deaf signing adults: the role of auditory cortex in functional reorganization following deafness. Neuropsychologia. 2022:166:108139. 10.1016/j.neuropsychologia.2021.108139. [DOI] [PubMed] [Google Scholar]
  2. Armstrong NM, Williams OA, Landman BA, Deal JA, Lin FR, Resnick SM. Association of poorer hearing with longitudinal change in cerebral white matter microstructure. JAMA Otolaryngol Head Neck Surg. 2020:146(11):1035–1042. 10.1001/jamaoto.2020.2497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Baguley DM, Chang P, Moffat DA. Tinnitus and vestibular schwannoma. Semin Hear. 2001:22(01):077–088. 10.1055/s-2001-13022. [DOI] [Google Scholar]
  4. Batalle D, Hughes EJ, Zhang H, Tournier J-D, Tusor N, Aljabar P, Wali L, Alexander DC, Hajnal JV, Nosarti C, et al.  Early development of structural networks and the impact of prematurity on brain connectivity. NeuroImage. 2017:149:379–392. 10.1016/j.neuroimage.2017.01.065. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Beer AL, Plank T, Meyer G, Greenlee MW. Combined diffusion-weighted and functional magnetic resonance imaging reveals a temporal-occipital network involved in auditory-visual object processing. Front Integr Neurosci. 2013:7:5. 10.3389/fnint.2013.00005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc. 1995:57(1):289–300. 10.1111/j.2517-6161.1995.tb02031.x. [DOI] [Google Scholar]
  7. Bolsterlee B, D’Souza A, Herbert RD. Reliability and robustness of muscle architecture measurements obtained using diffusion tensor imaging with anatomically constrained tractography. J Biomech. 2019:86:71–78. 10.1016/j.jbiomech.2019.01.043. [DOI] [PubMed] [Google Scholar]
  8. Bonna K, Finc K, Zimmermann M, Bola L, Mostowski P, Szul M, Rutkowski P, Duch W, Marchewka A, Jednoróg K, et al.  Early deafness leads to re-shaping of functional connectivity beyond the auditory cortex. Brain Imaging Behav. 2021:15(3):1469–1482. 10.1007/s11682-020-00346-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Brissenden JA, Somers DC. Cortico–cerebellar networks for visual attention and working memory. Curr Opin Psychol. 2019:29:239–247. 10.1016/j.copsyc.2019.05.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Calati R, Maller JJ, Meslin C, Lopez-Castroman J, Ritchie K, Courtet P, Artero S. Repatriation is associated with isthmus cingulate cortex reduction in community-dwelling elderly. World J Biol Psychiatry. 2018:19(6):421–430. 10.1080/15622975.2016.1258490. [DOI] [PubMed] [Google Scholar]
  11. Cardin V. Effects of aging and adult-onset hearing loss on cortical auditory regions. Front Neurosci. 2016:10:199. 10.3389/fnins.2016.00199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Chen J, Hu B, Qin P, Gao W, Liu C, Zi D, Ding X, Yu Y, Cui G, Lu L. Altered brain activity and functional connectivity in unilateral sudden sensorineural hearing loss. Neural Plast. 2020:2020:9460364. 10.1155/2020/9460364. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Choi JS, Wu F, Park S, Friedman RA, Kari E, Volker CCJ. Factors associated with unilateral hearing loss and impact on communication in US adults. Otolaryngol Head Neck Surg. 2021:165(6):868–875. 10.1177/0194599821995485. [DOI] [PubMed] [Google Scholar]
  14. Dell’Acqua F, Tournier J-D. Modelling white matter with spherical deconvolution: how and why?  NMR Biomed. 2019:32(4):e3945. 10.1002/nbm.3945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Desikan RS, Ségonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D, Buckner RL, Dale AM, Maguire RP, Hyman BT, et al.  An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. NeuroImage. 2006:31(3):968–980. 10.1016/j.neuroimage.2006.01.021. [DOI] [PubMed] [Google Scholar]
  16. Dhollander T, Clemente A, Singh M, Boonstra F, Civier O, Duque JD, Egorova N, Enticott P, Fuelscher I, Gajamange S, et al.  Fixel-based analysis of diffusion MRI: methods, applications, challenges and opportunities. NeuroImage. 2021:241:118417. 10.1016/j.neuroimage.2021.118417. [DOI] [PubMed] [Google Scholar]
  17. Falkenberg LE, Specht K, Westerhausen R. Attention and cognitive control networks assessed in a dichotic listening fMRI study. Brain Cogn. 2011:76(2):276–285. 10.1016/j.bandc.2011.02.006. [DOI] [PubMed] [Google Scholar]
  18. Finkl T, Hahne A, Friederici AD, Gerber J, Mürbe D, Anwander A. Language without speech: segregating distinct circuits in the human brain. Cereb Cortex. 2020:30(2):812–823. 10.1093/cercor/bhz128. [DOI] [PubMed] [Google Scholar]
  19. Foley RW, Shirazi S, Maweni RM, Walsh K, McConn Walsh R, Javadpour M, Rawluk D. Signs and symptoms of acoustic neuroma at initial presentation: an exploratory analysis. Cureus. 2017:9(11):e1846. 10.7759/cureus.1846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hackett TA. Information flow in the auditory cortical network. Hear Res. 2011:271(1–2):133–146. 10.1016/j.heares.2010.01.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Herlin B, Navarro V, Dupont S. The temporal pole: from anatomy to function—a literature appraisal. J Chem Neuroanat. 2021:113:101925. 10.1016/j.jchemneu.2021.101925. [DOI] [PubMed] [Google Scholar]
  22. Hickok G, Okada K, Serences JT. Area Spt in the human planum temporale supports sensory-motor integration for speech processing. J Neurophysiol. 2009:101(5):2725–2732. 10.1152/jn.91099.2008. [DOI] [PubMed] [Google Scholar]
  23. Horbruegger M, Loewe K, Kaufmann J, Wagner M, Schippling S, Pawlitzki M, Schoenfeld MA. Anatomically constrained tractography facilitates biologically plausible fiber reconstruction of the optic radiation in multiple sclerosis. NeuroImage Clinical. 2019:22:101740. 10.1016/j.nicl.2019.101740. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Horwitz B, Braun AR. Brain network interactions in auditory, visual and linguistic processing. Brain Lang. 2004:89(2):377–384. 10.1016/S0093-934X(03)00349-3. [DOI] [PubMed] [Google Scholar]
  25. Jeurissen B, Tournier J-D, Dhollander T, Connelly A, Sijbers J. Multi-tissue constrained spherical deconvolution for improved analysis of multi-shell diffusion MRI data. NeuroImage. 2014:103:411–426. 10.1016/j.neuroimage.2014.07.061. [DOI] [PubMed] [Google Scholar]
  26. Kamagata K, Zalesky A, Hatano T, Di Biase MA, El Samad O, Saiki S, Shimoji K, Kumamaru KK, Kamiya K, Hori M, et al.  Connectome analysis with diffusion MRI in idiopathic Parkinson’s disease: evaluation using multi-shell, multi-tissue, constrained spherical deconvolution. NeuroImage Clinical. 2018:17:518–529. 10.1016/j.nicl.2017.11.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kreutzer JS, DeLuca J, Caplan B. Encyclopedia of clinical neuropsychology. Springer New York; 2010. https://toc.library.ethz.ch/objects/pdf/z01_978-0-387-79947-6_01.pdf.
  28. Kuiper JJ, Lin Y-H, Young IM, Bai MY, Briggs RG, Tanglay O, Fonseka RD, Hormovas J, Dhanaraj V, Conner AK, et al.  A parcellation-based model of the auditory network. Hear Res. 2020:396:108078. 10.1016/j.heares.2020.108078. [DOI] [PubMed] [Google Scholar]
  29. Lazard DS, Collette J-L, Perrot X. Speech processing: from peripheral to hemispheric asymmetry of the auditory system. Laryngoscope. 2012:122(1):167–173. 10.1002/lary.22370. [DOI] [PubMed] [Google Scholar]
  30. Litwińczuk MC, Muhlert N, Cloutman L, Trujillo-Barreto N, Woollams A. Combination of structural and functional connectivity explains unique variation in specific domains of cognitive function. NeuroImage. 2022:262:119531. 10.1016/j.neuroimage.2022.119531. [DOI] [PubMed] [Google Scholar]
  31. Lomber SG, Meredith MA, Kral A. Cross-modal plasticity in specific auditory cortices underlies visual compensations in the deaf. Nat Neurosci. 2010:13(11):1421–1427. 10.1038/nn.2653. [DOI] [PubMed] [Google Scholar]
  32. Lynn CW, Bassett DS. The physics of brain network structure, function and control. Nature Reviews Physics. 2019:1(5):318–332. 10.1038/s42254-019-0040-8. [DOI] [Google Scholar]
  33. McLaren ME, Szymkowicz SM, O’Shea A, Woods AJ, Anton SD, Dotson VM. Dimensions of depressive symptoms and cingulate volumes in older adults. Transl Psychiatry. 2016:6(4):e788. 10.1038/tp.2016.49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mišić B, Betzel RF, Griffa A, de  ReusMA, He Y, Zuo X-N, van den  HeuvelMP, Hagmann P, Sporns O, Zatorre RJ. Network-based asymmetry of the human auditory system. Cereb Cortex. 2018:28(7):2655–2664. 10.1093/cercor/bhy101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Moffat DA, Baguley DM, Beynon GJ, Da Cruz M. Clinical acumen and vestibular schwannoma. Am J Otol. 1998:19(1):82–87  https://www.ncbi.nlm.nih.gov/pubmed/9455955. [PubMed] [Google Scholar]
  36. Palejwala AH, O’Connor KP, Milton CK, Anderson C, Pelargos P, Briggs RG, Conner AK, O’Donoghue DL, Glenn CA, Sughrue ME. Anatomy and white matter connections of the fusiform gyrus. Sci Rep. 2020:10(1):13489. 10.1038/s41598-020-70410-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ponticorvo S, Manara R, Cassandro E, Canna A, Scarpa A, Troisi D, Cassandro C, Cuoco S, Cappiello A, Pellecchia MT, et al.  Cross-modal connectivity effects in age-related hearing loss. Neurobiol Aging. 2022:111:1–13. 10.1016/j.neurobiolaging.2021.09.024. [DOI] [PubMed] [Google Scholar]
  38. Price CJ, Crinion JT, Macsweeney M. A generative model of speech production in Broca’s and Wernicke’s areas. Front Psychol. 2011:2:237. 10.3389/fpsyg.2011.00237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Qi R, Su L, Zou L, Yang J, Zheng S. Altered gray matter volume and white matter integrity in sensorineural hearing loss patients: a VBM and TBSS study. Otol Neurotol. 2019:40(6):e569–e574. 10.1097/MAO.0000000000002273. [DOI] [PubMed] [Google Scholar]
  40. R Core Team.  R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2013. https://www.R-project.org/.
  41. Robinson FR, Fuchs AF. The role of the cerebellum in voluntary eye movements. Annu Rev Neurosci. 2001:24(1):981–1004. 10.1146/annurev.neuro.24.1.981. [DOI] [PubMed] [Google Scholar]
  42. Rönnberg J, Danielsson H, Rudner M, Arlinger S, Sternäng O, Wahlin A, Nilsson L-G. Hearing loss is negatively related to episodic and semantic long-term memory but not to short-term memory. J Speech Lang Hear Res. 2011:54(2):705–726. 10.1044/1092-4388(2010/09-0088). [DOI] [PubMed] [Google Scholar]
  43. Saur D, Schelter B, Schnell S, Kratochvil D, Küpper H, Kellmeyer P, Kümmerer D, Klöppel S, Glauche V, Lange R, et al.  Combining functional and anatomical connectivity reveals brain networks for auditory language comprehension. NeuroImage. 2010:49(4):3187–3197. 10.1016/j.neuroimage.2009.11.009. [DOI] [PubMed] [Google Scholar]
  44. Schlauch RS, Anderson ES, Micheyl C. A demonstration of improved precision of word recognition scores. J Speech Lang Hear Res. 2014:57(2):543–555. 10.1044/2014_JSLHR-H-13-0017. [DOI] [PubMed] [Google Scholar]
  45. Schmidt SA, Akrofi K, Carpenter-Thompson JR, Husain FT. Default mode, dorsal attention and auditory resting state networks exhibit differential functional connectivity in tinnitus and hearing loss. PLoS One. 2013:8(10):e76488. 10.1371/journal.pone.0076488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Sethi SS, Zerbi V, Wenderoth N, Fornito A, Fulcher BD. Structural connectome topology relates to regional BOLD signal dynamics in the mouse brain. Chaos. 2017:27(4):047405. 10.1063/1.4979281. [DOI] [PubMed] [Google Scholar]
  47. Shang Y, Hinkley LB, Cai C, Mizuiri D, Cheung SW, Nagarajan SS. Cross-modal plasticity in adult single-sided deafness revealed by alpha band resting-state functional connectivity. NeuroImage. 2020:207:116376. 10.1016/j.neuroimage.2019.116376. [DOI] [PubMed] [Google Scholar]
  48. Smith RE, Tournier J-D, Calamante F, Connelly A. Anatomically-constrained tractography: improved diffusion MRI streamlines tractography through effective use of anatomical information. NeuroImage. 2012:62(3):1924–1938. 10.1016/j.neuroimage.2012.06.005. [DOI] [PubMed] [Google Scholar]
  49. Sporns O. Contributions and challenges for network models in cognitive neuroscience. Nat Neurosci. 2014:17(5):652–660. 10.1038/nn.3690. [DOI] [PubMed] [Google Scholar]
  50. Wei Z, Fan Z, Qi Z, Tong Y, Guo Q, Chen L. Reorganization of auditory-visual network interactions in long-term unilateral postlingual hearing loss. J Clin Neurosci. 2021:87:97–102. 10.1016/j.jocn.2021.02.017. [DOI] [PubMed] [Google Scholar]
  51. Xu H, Fan W, Zhao X, Li J, Zhang W, Lei P, Liu Y, Wang H, Cheng H, Shi H. Disrupted functional brain connectome in unilateral sudden sensorineural hearing loss. Hear Res. 2016:335:138–148. 10.1016/j.heares.2016.02.016. [DOI] [PubMed] [Google Scholar]
  52. Yang M, Chen H-J, Liu B, Huang Z-C, Feng Y, Li J, Chen J-Y, Zhang L-L, Ji H, Feng X, et al.  Brain structural and functional alterations in patients with unilateral hearing loss. Hear Res. 2014:316:37–43. 10.1016/j.heares.2014.07.006. [DOI] [PubMed] [Google Scholar]
  53. Zhang J, Wang J, Wu Q, Kuang W, Huang X, He Y, Gong Q. Disrupted brain connectivity networks in drug-naive, first-episode major depressive disorder. Biol Psychiatry. 2011:70(4):334–342. 10.1016/j.biopsych.2011.05.018. [DOI] [PubMed] [Google Scholar]
  54. Zhang G-Y, Yang M, Liu B, Huang Z-C, Li J, Chen J-Y, Chen H, Zhang P-P, Liu L-J, Wang J, et al.  Changes of the directional brain networks related with brain plasticity in patients with long-term unilateral sensorineural hearing loss. Neuroscience. 2016:313:149–161. 10.1016/j.neuroscience.2015.11.042. [DOI] [PubMed] [Google Scholar]
  55. Zhang Y, Mao Z, Feng S, Liu X, Lan L, Zhang J, Yu X. Altered functional networks in long-term unilateral hearing loss: a connectome analysis. Brain Behav. 2018a:8(2):e00912. 10.1002/brb3.912. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhang Y, Mao Z, Feng S, Liu X, Zhang J, Yu X. Monaural-driven functional changes within and beyond the auditory cortical network: evidence from long-term unilateral hearing impairment. Neuroscience. 2018b:371:296–308. 10.1016/j.neuroscience.2017.12.015. [DOI] [PubMed] [Google Scholar]
  57. Zhao T, Cao M, Niu H, Zuo X-N, Evans A, He Y, Dong Q, Shu N. Age-related changes in the topological organization of the white matter structural connectome across the human lifespan. Hum Brain Mapp. 2015:36(10):3777–3792. 10.1002/hbm.22877. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Zou Y, Ma H, Liu B, Li D, Liu D, Wang X, Wang S, Fan W, Han P. Disrupted topological organization in white matter networks in unilateral sudden sensorineural hearing loss. Front Neurosci. 2021:15:666651. 10.3389/fnins.2021.666651. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplementarymaterial_bhae220
supplementarymaterial_bhae220.docx (4.6MB, docx)

Data Availability Statement

Anonymized data and code used for this study is available upon reasonable request from the corresponding author.

Articles from Cerebral Cortex (New York, NY) are provided here courtesy of Oxford University Press

RESOURCES