My research into vestibular perception began with a simple clinical observation. An elderly patient with cerebral small-vessel disease had preserved vestibular reflexes but appeared strikingly unable to perceive vestibularly driven self-motion. This raised a fundamental question: can the brain process vestibular information sufficiently to generate normal reflexes, yet fail to generate the conscious sensation of movement?
During my PhD, I developed methods to address this experimentally. In Seemungal et al. (2004), we measured vestibular perceptual and vestibulo-ocular reflex thresholds simultaneously during whole-body rotation. This provided a quantitative means of separating the brain mechanisms responsible for the sensation of self-motion from those producing vestibular reflexes, and established the experimental framework for much of my subsequent work.
A second question was whether sensing movement and knowing where that movement has taken us are also separable processes. Using repetitive transcranial magnetic stimulation, we showed that perturbing posterior parietal cortex disrupted vestibular-guided spatial orientation or path integration, while leaving basic self-motion perception relatively intact (Seemungal et al., 2008). In simple terms, the cortical computation answering “Where am I? could be disturbed without abolishing the more elementary sensation Am I moving?
During my subsequent Academy of Medical Sciences Fellowship, I moved from characterising these perceptual dissociations towards identifying their underlying neural architecture.
Our focal lesion study provided an important test of the cortical findings. In patients with acute hemispheric lesions, damage involving temporoparietal regions impaired vestibular-guided estimates of travelled distance and motion duration, yet self-motion perception itself remained normal (Kaski et al., Brain, 2016).
This was consistent with our structural imaging work, particularly Nigmatullina et al. (Cerebral Cortex, 2015), which showed that vestibular self-motion perception relates not simply to a single cortical “vestibular area”, but to an extensive cerebral white-matter network.
Together, these studies suggested an important distinction: spatial orientation can depend critically upon focal temporoparietal mechanisms, whereas conscious vestibular self-motion sensation appears to depend upon a more distributed, and probably bihemispheric, brain network.
Clinical observations had also suggested that the dissociation between vestibular perception and reflex function begins below the cerebral cortex.
In my review Neuro-otological emergencies, I described the striking presentation of acute focal cerebellar stroke producing severe vertigo despite the absence of the vestibulo-ocular reflex nystagmus that might normally accompany such a sensation. This clinical observation was subsequently supported by Shaikh and colleagues, who described patients with posterior cerebellar infarction experiencing persistent vestibular motion despite little or no accompanying spontaneous nystagmus.
A different form of perceptuo-reflex dissociation emerged from a longitudinal single-case study of Miller Fisher syndrome. During recovery, the time course of the patient’s symptoms more closely followed recovery of vestibular motion perception than vestibulo-ocular reflex function, suggesting that perceptual and ocular responses to the same vestibular disturbance could recover independently.
These clinical observations helped motivate our study of adaptive vestibular plasticity in highly trained ballet dancers (Nigmatullina et al., Cerebral Cortex, 2015). Repeated exposure to rotation had produced a striking uncoupling between vestibular self-motion perception and the VOR. Crucially, the anatomical correlate of this dissociation involved the vestibular cerebellum, suggesting that cerebellar processing can differentially regulate vestibular signals contributing to conscious perception and those driving reflex function.
The same study showed that robust vestibular perception was associated with an extensive cerebral white-matter network, linking this subcortical processing to the distributed cortical system required for conscious self-motion.
We then asked whether other subcortical systems could actively regulate this perceptual signal. In Yousif et al. (2016), studying patients with Parkinson’s disease undergoing deep-brain stimulation of the pedunculopontine nucleus (PPN), we showed that PPN stimulation lowered vestibular perceptual thresholds and altered the relationship between vestibular perception and postural control. This provided evidence that ascending brainstem systems can regulate the gain of vestibular signals relevant both to conscious perception and to balance.
Since taking over leadership of the Centre for Vestibular Neurology from my mentor, Professor Adolfo Bronstein, in 2020, my research programme has increasingly focused on how these mechanisms present clinically, how they affect recovery and balance, and whether they can ultimately be therapeutically modified.
Traumatic brain injury provided a powerful model in which to test the distributed-network hypothesis. In Calzolari et al. (Brain, 2021), we identified a syndrome we termed vestibular agnosia: markedly impaired or absent conscious perception of self-motion despite preserved peripheral vestibular and vestibulo-ocular reflex function.
In that study, we showed that vestibular agnosia is common after acute traumatic brain injury and is associated with impaired balance and disruption of cerebral white-matter pathways. It therefore provides a clinical counterpart to the earlier experimental dissociations between vestibular sensation, reflexes and spatial processing.
This work was extended by Hadi et al. (2022), in which functional connectivity analysis showed that vestibular agnosia is associated with disruption of both intra- and inter-hemispheric networks. This helps explain why relatively circumscribed unilateral cortical lesions can spare basic self-motion perception, whereas diffuse brain injury can profoundly impair it.
Our subsequent longitudinal work (Hadi et al., 2025) showed that recovery of vestibular perception and recovery of balance are closely related, providing further evidence that these perceptual mechanisms are functionally important rather than simply an epiphenomenon of brain injury.
Having identified the anatomical networks associated with vestibular perception, we have also begun asking how the emergence of self-motion is represented dynamically in human brain activity.
In Hadi et al. (2024), we used EEG to separate cortical activity related to vestibular sensory activation from activity associated with the conscious sensation of movement. The results suggested a particularly interesting dissociation: alpha-band activity was associated with vestibular sensory-signal processing, whereas theta-band activity tracked vestibular-mediated self-motion perception. Importantly, self-motion could be studied both when generated by actual vestibular stimulation and when the sensation was elicited without corresponding bottom-up vestibular activation.
In subsequent work (Hadi et al., 2026), we extended the EEG approach using natural whole-body vestibular stimulation in healthy participants and patients with bilateral vestibular dysfunction. Vestibular stimulation evokes measurable cortical potentials and changes in oscillatory activity, with theta activity closely related to the dynamics of motion stimulation.
These studies provide a potential electrophysiological bridge between the anatomical network and the percept itself. They raise the possibility that vestibular sensory information is transformed through distinct oscillatory processes, with theta activity contributing to the neural dynamics through which vestibular information becomes conscious self-motion. An objective EEG signature of this process could ultimately also provide a means of studying vestibular perception in patients who cannot reliably report what they experience.
More recently, we have extended the subcortical work from the cerebellum and brainstem towards the thalamus.
We exploited MRI-guided focused-ultrasound thalamotomy as a unique human model for probing these pathways. In Ciocca et al. (2024), focal sonication within cerebellar-thalamic regions produced powerful illusions of whole-body self-motion, particularly when targeting the posterior subthalamic region. These observations provide direct human evidence that cerebellar-thalamic pathways can modulate conscious vestibular sensation.
Our current work extends this model further. Findings presently under review link individual differences in vestibular self-motion perception to cerebellothalamic circuitry and dorsomedial-thalamic-prefrontal connectivity.
Together, these findings support a model in which conscious vestibular perception emerges through an interaction between subcortical gating of vestibular information, temporally organised neural activity and distributed cortical integration.
An important theme running through this work is that vestibular perception is not simply responsible for the subjective experience of vertigo. It is closely linked to postural control, navigation and the monitoring of our own movement through the world.
This raises a wider evolutionary question. The increasing importance of higher-order vestibular processing may have accompanied the evolution of habitual bipedal stance and gait, where maintaining balance had to coexist with prehensile upper-limb activity, navigation and increasingly complex cognitive behaviour.
Conscious access to self-motion information may therefore form part of a broader system allowing the brain to monitor, predict and modify balance while simultaneously performing higher-order cognitive and motor tasks.
Vestibular agnosia has an immediate clinical consequence: absence of dizziness does not necessarily mean absence of vestibular disease.
In our acute TBI work, vestibular agnosia produced approximately a sevenfold reduction in clinical recognition of benign paroxysmal positional vertigo (BPPV), because patients with physiologically active BPPV did not necessarily experience or report vertigo (Calzolari et al., 2021).
There is an additional and clinically important interaction between the mechanism producing BPPV and vestibular agnosia. In a prospective study led by Heiko Rust (Rust et al., 2022), we showed that post-traumatic BPPV is force-dependent: the likelihood of BPPV increases with the magnitude of biomechanical forces acting on the head, with BPPV particularly common in patients with skull fracture.
This creates a concerning clinical convergence. Patients exposed to substantial traumatic forces are especially likely to develop BPPV, while patients with central disruption of vestibular perception may fail to experience or communicate the dizziness that would normally lead clinicians to diagnose it. Thus, patients with vestibular agnosia may be among those most likely to harbour post-traumatic BPPV while simultaneously being among the least likely to have it recognised and treated.
Subsequent work by Harrell and colleagues has reinforced the wider problem, showing that systematic examination identifies dramatically more BPPV than symptom-led screening in brain-injured populations.
The vestibular agnosia work has also directly informed subsequent clinical studies at the Centre for Vestibular Neurology. Rebecca Smith, a colleague whom I mentored at the Centre and now an NIHR Postdoctoral Fellow, completed an NIHR-funded prospective study of the treatment of BPPV following acute traumatic brain injury. The design of that study was informed by our finding that vestibular agnosia can render BPPV clinically “silent”, meaning that reliance on dizziness symptoms alone risks missing an important and readily treatable vestibular disorder.
My collaborators at the Centre for Vestibular Neurology have also taken these ideas into a large population of otherwise healthy older adults. Work led by Li, and Ellmers shows that impaired vestibular perception also occurs in healthy ageing and is associated with poorer balance, suggesting that vestibular agnosia may represent a more general mechanism contributing to age-related instability.
The translational goal is therefore twofold: first, to develop clinical approaches that identify vestibular disorders even when the expected symptoms are absent; and second, to determine whether the neural mechanisms underlying vestibular perception can themselves be modified.
Our findings with PPN deep-brain stimulation, together with ongoing work on cerebellar-thalamic circuitry, electrophysiological mechanisms and pharmacological modulation, suggest that vestibular perceptual processing is not fixed.
Ultimately, understanding how the brain transforms vestibular signals into conscious self-motion may provide new ways of diagnosing—and potentially treating—disorders of balance across traumatic brain injury, neurodegeneration and ageing.