Acute bilateral central deafness following cardioembolic ischemic stroke: insights from EEG spectral and connectivity analysis

Abstract

Stroke is a leading cause of adult disability, with outcomes depending on the lesion’s location and severity. While motor and language deficits are common, rare manifestations like central auditory dysfunction can be equally debilitating. This case report presents a 57-year-old woman who developed acute bilateral central deafness following a cardioembolic ischemic stroke. To investigate underlying mechanisms, advanced EEG analyses were conducted, including spectral power analysis and eLORETA-based functional connectivity, across key frequency bands. The Small World index was applied to evaluate changes in brain network organization over 6 weeks. Results revealed disrupted connectivity in bilateral auditory and associative cortices, with significant spectral power alterations. Alpha and gamma bands showed distinct recovery dynamics, while connectivity metrics indicated a shift toward global integration, possibly compensating for local disconnections. These findings demonstrate the utility of EEG biomarkers in monitoring post-stroke neural reorganization and support their potential role in guiding rehabilitation for sensory deficits.

Introduction

Stroke can cause various neurological deficits, depending on the location and extent of the brain injury. While many strokes result in motor or language impairments, some cases lead to rare but equally debilitating effects, such as central auditory dysfunction. Since the auditory input is processed in both auditory cortices via bilateral pathways, central deafness is rarely observed after stroke.

A systematic review published in 2021 identified only 44 reported cases of stroke-associated cortical deafness, all involving bilateral lesions of the auditory cortices—predominantly within the Heschl’s gyrus and temporoparietal junctions (~ 30% of cases). In this report, just 13% of cases involved subcortical hemispheric lesions, and only ~ 0.05% involved brainstem infarcts [1]. Other reports link auditory deficits to infarcts in the vertebrobasilar circulation, particularly in the AICA or PICA territories, although such cases are usually unilateral and subcortical, rarely leading to complete cortical deafness [2].

Although the exact incidence of cortical deafness is not known, broader data suggest that auditory-processing deficits are under-recognized after stroke. A case–control study reported that up to 40% of younger stroke survivors (18–60 years) exhibit central or peripheral hearing impairment when specifically tested, despite limited clinical reporting [3]. This aligns with the notion that central auditory dysfunction is likely underestimated in routine post-stroke evaluations, where auditory deficits are rarely investigated unless prompted by specific neurological symptoms or complaints [4, 5].

This under-recognition highlights the need for a comprehensive diagnostic approach, combining detailed clinical assessments with advanced neurophysiological techniques, such as auditory brainstem responses and cortical evoked potentials, to understand the underlying mechanisms of sensory disruption and guide targeted rehabilitation strategies [5]. Indeed, the clinical trajectory is highly variable. In some patients, cortical deafness is transient, evolving over weeks to months toward less severe auditory syndromes, such as auditory agnosia or selective word-deafness. In others, it remains persistent, particularly when extensive bilateral cortical damage is present, sometimes accompanied by disrupted frontotemporal connectivity affecting awareness networks [6].

In this context, electroencephalography (EEG) has emerged as an invaluable non-invasive tool for monitoring the brain’s electrical activity. One of the most established and widely used approaches in EEG analysis is the examination of power spectral density (PSD), which quantifies signal power across frequency bands, revealing cortical dysfunction through increased delta and theta (indicating cortical dysfunction and disrupted connectivity) and decreased alpha and beta (reflecting sensory and cognitive deficits). Additionally, multi-channel EEG enhances topographical resolution, helping to map ischemic injury locations [7].

In addition to PSD analysis, a more recent but widely used technique for investigating brain activity is graph theory, which offers insights into functional connectivity, modeling brain regions as nodes and their interactions as edges [8]. The Small World (SW) index measures network efficiency where changes post-stroke indicate functional reorganization: a reduced SW index in low-frequency bands suggests disrupted global connectivity, while alpha/beta variations may reflect recovery or persistent connectivity deficits.

This case report describes a 57-year-old woman with acute bilateral central deafness after a cardioembolic ischemic stroke. EEG analyses were conducted at three post-stroke time points (T0, T1, T2) over 6 weeks, assessing spectral power and functional connectivity via SW index to examine brain network changes.

Methods

A 57-year-old woman developed acute bilateral deafness after a stress-related cardioembolic ischemic stroke. Initial symptoms included dyspnea and chest pain, revealing reduced ejection fraction (35%) and a suspected apical thrombus, treated with enoxaparin and antiplatelet therapy.

Two days later, she showed left-sided hemiparesis; imaging confirmed ischemic stroke with additional lesions and hemorrhagic infiltration. She developed severe bilateral hypoacusis, non-fluent aphasia (NIHSS = 7), and left sensory extinction. Cardiac imaging indicated thrombus resolution and improved ejection fraction (70%), suggesting Takotsubo cardiomyopathy, and intensive neurorehabilitation was initiated. EEG recordings were conducted at rest with eyes closed, using a 32-channel system. The patient underwent three EEG sessions at 14 days (T0), 3 weeks (T1), and 6 weeks (T2) post-stroke. Data preprocessing included filtering, ICA-based artifact removal, and epoch segmentation [9].

Power spectral density (PSD)

The power spectral density (PSD) provides insight on how signal power is distributed across specific frequency bands. PSD was calculated using MATLAB’s pwelch function. The frequency bands examined included delta (2–4 Hz), theta (4–8 Hz), alpha 1 (8–11 Hz), alpha 2 (11–13 Hz), beta 1 (13–20 Hz), beta 2 (20–30 Hz), and gamma (30–45 Hz). Topographical maps were created to visualize spectral power distribution across the scalp, with a color scale from blue (low power) to red (high power). To focus on auditory processing, spectral values were averaged across the temporal region (F7, T7, P7, F8, T8, P8), allowing for a more targeted analysis of regional spectral power changes over time and conditions.

Functional connectivity analysis

Functional connectivity was assessed using eLORETA, an EEG method estimating cortical current density in 6239 voxels (5-mm resolution) within a realistic head model. Functional connectivity was computed across 84 ROIs (42 Bas per hemisphere). In addition, specific analysis was conducted on the primary auditory cortex (acoustic network – BA 13, 20, 21, 22, 41, 42 in each hemisphere, resulting in a total of 12 ROIs), analyzing frequency-specific connections via Intracortical Lagged Linear Coherence [9].

Graph analysis

Brain networks were modeled using graph theory, with nodes as ROIs and edges defined by coherence. An undirected, weighted network was created, and the Small World (SW) index was used to evaluate network efficiency across frequency bands [10].

Results

Power spectral density (PSD)

To investigate the modulation of brain activity over time, power spectral density (PSD) was calculated across the main EEG frequency bands, and its topographic distribution on the scalp was assessed. The findings revealed temporal dynamics in the alpha 1, alpha 2, and gamma bands (Fig. 1).

Fig. 1.

Topographic distributions of the PSD in the eyes-closed condition. PSD values are represented by a color bar ranging from blue (indicating lower PSD values) to red (indicating higher PSD values). The maps illustrate the spatial distribution of PSD across the scalp for three time points (T0, T1, T2) in the alpha 1, alpha 2, and gamma bands

The alpha 1 and alpha 2 bands showed a progressive increase in PSD over time, with activity shifting from low levels at T0 to marked rises by T2, especially in the posterior and occipital regions. In contrast, the gamma band exhibited a consistent decrease in PSD, starting high at T0 and substantially declining by T2, indicating reduced high-frequency cortical power over time.

Focusing on the temporal ROI (Fig. 2), PSD values decreased consistently across all frequency bands over time. At T0, spectral power was highest, diminishing to intermediate levels at T1 and reaching the lowest values at T2. The most significant reductions were observed in the delta, theta, alpha 1, and gamma bands, highlighting a progressive decline in cortical power. Despite similar curve shapes across time points, the T2 curve showed a marked reduction compared to the others.

Fig. 2.

Power spectral density (PSD) trends in the temporal ROI under eyes-closed condition across three time points (T0, T1, and T2) and in the main EEG frequency bands (delta, theta, alpha 1, alpha 2, beta 1, beta 2, and gamma). The blue line represents the PSD values at T0, the orange line represents T1, and the green line represents T2

Small world

The Small World (SW) index was calculated across the main frequency bands for three time points (T0, T1, T2) to assess network organization under the eyes-closed condition (Fig. 3A).

Fig. 3.

A Small World trends under eyes-closed condition across three time points (T0, T1, and T2) and in the main EEG frequency bands (delta, theta, alpha 1, alpha 2, beta 1, beta 2, and gamma). The blue line represents the SW values at T0, the orange line represents T1, and the green line represents T2. B Small World trends under eyes-closed condition across three time points (T0, T1, and T2) and in the main EEG frequency bands (delta, theta, alpha 1, alpha 2, beta 1, beta 2, and gamma) in the acoustic network. The blue line represents the SW values at T0, the orange line represents T1, and the green line represents T2

In the delta band, values remained stable, while the theta band steadily increased over time. In the alpha 1 band, later measurements showed significantly higher values, whereas the alpha 2 band started lower and then increased to similar levels. The beta 1 band stayed constant, and the beta 2 and gamma bands were highest initially before decreasing.

Focusing on the acoustic network (Fig. 3B), the trends were similar. The delta band peaked in the middle session, the theta band was highest early on, and the alpha 1 band reached its lowest mid-session. The alpha 2 band increased from the initial measurement, the beta 1 band gradually rose, and the beta 2 and gamma bands were highest initially and then declined.

Discussion

The present study provides an examination of the time dynamics of cortical activity and network organization in a patient with central deafness following stroke, revealing frequency-specific modulations and the underlying pathophysiological mechanisms involved in stroke recovery.

The progressive increase in PSD within the alpha 1 and alpha 2 bands over time suggests a reorganization of cortical inhibition and information processing within posterior regions. The enhancement of alpha power, particularly in the posterior regions, may reflect an adaptive mechanism aimed at re-establishing functional balance following ischemic insult [11]. Conversely, the consistent reduction in gamma band power suggests a loss of high-frequency cortical activity that is often associated with local neural synchrony and the integration of perceptual information [12].

A more focused analysis of the temporal region revealed a pronounced decrease in PSD across all frequency bands under eyes-closed conditions. This observation supports the notion of sensory deprivation leading to neural loss/fatigue and diminished cortical responsiveness. The marked divergence of the T2 PSD curve in the temporal ROI suggests that prolonged sensory disconnection might trigger a reorganization process, potentially as the brain attempts to compensate for auditory input cortical deafferentation. Similar reductions in spectral power have been noted in studies investigating cortical reorganization after sensory loss [13], underscoring the vulnerability of these regions to ischemic damage.

Graph theoretical analysis further enriched our understanding of post-stroke network dynamics. The Small World (SW) index demonstrated distinct changes across frequency bands. In the delta band, the relative stability of the SW index, with a slight increase at T2, might suggest that the low-frequency networks maintain some degree of functional integrity, as has been observed in other stroke populations [8]. The progressive increase in the theta band SW index indicates a move toward less ordered network configurations, a phenomenon that could represent compensatory reorganization aimed at supporting cognitive integration. The sharp decline in the SW index in the alpha 1 band at T2, contrasted by a recovery trend in the alpha 2 band, might reflect the complex interplay between disrupted connectivity and neural plasticity.

In the acoustic network, the gradual decrease in the gamma band SW index suggests that high-frequency network connectivity is particularly susceptible to ischemic damage, potentially contributing to the clinical manifestation of central deafness. In contrast, changes in the beta bands are consistent with previous studies that have reported compensatory network reconfigurations aimed at restoring efficient auditory processing after stroke [14]. This reorganization is in line with the notion that network plasticity, even in the face of significant neural injury, may underlie functional recovery processes [5, 9].

The findings suggest that EEG spectral and connectivity markers may serve as valuable, non-invasive tools for monitoring neural recovery and guiding personalized rehabilitation in central deafness following stroke. However, further research with larger cohorts and multimodal imaging is needed to validate these findings and understand the underlying mechanisms.

Author contribution

AC: original idea, writing—original draft; data curation; investigation

FM: investigation; writing—reviewing and editing

GS: writing—reviewing and editing

FI: writing—reviewing and editing

AM: writing—reviewing and editing

CP: data curation; reviewing and editing

PV: data curation; reviewing and editing

PMR: writing—reviewing and editing

FV: original idea; investigation; writing—reviewing and editing

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval

Approval was obtained from the ethics committee of IRCCS San Raffaele Roma. The procedures used in this study adhere to the tenets of the Declaration of Helsinki.

Consent to participate

Informed consent was obtained from the participant.

Consent to publish

The participant has consented to the submission of the case report to the journal.

Competing interests

The authors declare no competing interests.