14The Evidence — Neuroimaging and Electrophysiology

FWhy imaging changed the field

For most of its history tinnitus could only be inferred — from animal recordings, from patient report, from response to treatment. Human neuroimaging and electrophysiology changed that by letting us watch the living tinnitus brain. The headline finding is consistent across modalities: tinnitus is accompanied by altered activity and connectivity not just in the auditory system but in limbic, attentional and default-mode networks, supporting the view of tinnitus as a distributed central disorder [2009].

A landmark synthesis by Elgoyhen, Langguth and colleagues reviewed the human imaging evidence and argued that no single brain region ‘is’ tinnitus; rather, the percept and its distress emerge from interacting networks, and the search for a clean biomarker must reckon with that complexity [2015].

TfMRI and PET: resting hyperactivity and limbic coupling

Early PET work by Lockwood and colleagues showed increased cerebral blood flow in auditory cortex during tinnitus, and crucially demonstrated limbic-system involvement and evidence of neural plasticity — a foundational result for the network view [1998]. Functional MRI extended this, revealing elevated resting activity in Heschl’s gyrus and superior temporal gyrus and, in distressed patients, enhanced functional connectivity between auditory cortex and limbic structures such as the amygdala and parahippocampus [2011].

Resting-state studies then implicated the large-scale networks: tinnitus and hearing loss differentially alter the default-mode, dorsal-attention and auditory resting-state networks, suggesting the phantom intrudes via dysfunctional attention and self-monitoring systems rather than auditory cortex alone [2013].

CEEG/MEG: the oscillatory signature

Electromagnetic methods add millisecond resolution and reveal a characteristic spectral signature. Using MEG, Weisz and colleagues showed that tinnitus is associated with abnormal spontaneous activity — reduced alpha (8–12 Hz), reflecting lost inhibitory tone, together with enhanced delta/theta in deafferented regions and increased gamma (>30 Hz) that tracks tinnitus loudness and distress [2005]. This delta-gamma coupling is the human fingerprint of thalamocortical dysrhythmia [1999].

Source-localised EEG (for example LORETA) places these abnormalities not only in primary and secondary auditory cortex but also in prefrontal, parietal and limbic generators — converging with fMRI on the multi-network hypothesis and offering candidate targets for neuromodulation [2009].

CStructural imaging and evoked potentials

Structural MRI with voxel-based morphometry and diffusion tensor imaging adds a slower-changing layer of evidence. Landgrebe and colleagues reported grey-matter decreases in both auditory and non-auditory regions in tinnitus [2009], while Husain and colleagues used combined VBM and DTI to disentangle the structural changes attributable to hearing loss from those linked to tinnitus itself — an essential methodological step [2011].

At the periphery-to-brainstem end, evoked potentials are informative: reduced ABR wave-I amplitude in patients with normal audiograms is taken as physiological evidence of cochlear synaptopathy — hidden hearing loss — underlying tinnitus, the bridge between peripheral trigger and central consequence [2011].

CThe biomarker problem: confounds and heterogeneity

Despite this convergence, no validated objective biomarker of tinnitus exists. Two problems dominate. First, the hearing-loss confound: most tinnitus patients have some hearing loss, which itself reshapes the auditory brain, so studies that fail to match controls for audiometric status may attribute hearing-loss effects to tinnitus — the very reason combined-modality, carefully-matched designs were developed [2011]. Second, heterogeneity: tinnitus is many disorders wearing one name, and patients differ in laterality, pitch, distress and aetiology, blurring group averages [2015].

The pragmatic conclusion is that imaging has decisively established where and how tinnitus lives in the brain and has generated rational neuromodulation targets, but a single, clinic-ready diagnostic scan remains aspirational. Progress will come from large, harmonised, well-phenotyped cohorts rather than small case-control studies [2009].