Imaging in vertigo — when, what, and why.

Vertigo accounts for a meaningful share of dizziness visits to emergency departments and clinics, and most of it is peripheral and benign.1Yet a small, dangerous minority — cerebellar infarct, vertebral dissection, a vestibular schwannoma — looks indistinguishable from inner-ear vertigo at first pass. Imaging is the safety net.

The job of the imaging chapter is not to scan everyone. It is to recognise the clinical situations where imaging changes management, choose the modality that answers the clinical question, and read the resulting study with enough vocabulary to know whether the answer was found.

The TiTrATE framework — Timing, Triggers and Targeted Examination — points the clinician toward the imaging question well before the modality decision.2 Acute vestibular syndrome with a central HINTS pattern needs DWI MRI; a typical t-EVS with a positive Dix-Hallpike needs no imaging. The imaging choice is a consequence of the bedside hypothesis.

In emergency settings the prevalence of posterior-circulation stroke among dizzy patients is small but non-trivial — about one in twenty in unselected ED cohorts.3 The challenge is that early-window DWI is far from sensitive: HINTS performed by trained clinicians outperforms early MRI for the first 24–48 h.4Imaging supports — it does not replace — bedside reasoning.

The clinically useful taxonomy of vertigo imaging maps onto three questions: is there a posterior-circulation stroke? Is there a retrocochlear lesion? Is there a third-window or other bony anomaly? Each question maps to a different modality — DWI MRI ± CTA, gadolinium MRI of the IAC, HRCT in the Pöschl plane — and each modality has a characteristic false-negative profile that drives follow-up.

The remaining material in this chapter walks the modalities in turn (MRI, CT, CTA/MRA, fMRI, PET), then the disorder-by-modality matrix, the practical balance of advantages and limitations, and the AI-augmented future. Use the decision tree below as the menu.