Mapping the vestibular cerebellar evoked potential (VsCEP) following air- and bone-conducted vestibular stimulation

Vestibular cerebellar evoked potentials (VsCEPs) were recorded from over the occipital and cerebellar regions of the scalp using bone-conducted (BC) stimuli applied at the mastoids (impulsive accelerations and 500 Hz) and 500 Hz acoustic tones (AC). Ten healthy subjects were tested. Electrodes were positioned over the midline (Oz, Iz, CBz) and at 3, 6 and 9 cm intervals lateral to the midline electrodes bilaterally. Additional electrodes were also positioned over posterior neck muscles (SPL1 and SPL2). The largest evoked potentials on average were recorded from the electrodes 3 and 6 cm lateral to the Iz and CBz midline locations. BC stimuli produced short latency potentials on the side contralateral to the stimulated mastoid and were dependent on stimulus polarity. Positive polarity stimuli produced biphasic VsCEPs at approximately 12 and 17 ms (P12-N17) for BC impulses and 10 and 15 ms (P10-N15) for BC 500 Hz stimuli. Following the initial excitation, there was a period of suppression of background activity lasting an average of 16.8 ms for positive polarity BC impulses. Negative polarity stimuli produced later VsCEPs both for BC impulses (P20-N26) and BC 500 Hz (P13-N18). VsCEPs to AC 500 Hz stimuli lateralised to the contralateral side and were larger for right than left ear stimulation. Stimulus polarity (condensation and rarefaction) did not alter the timing of the VsCEPs to AC 500 Hz tones. No evoked response was recorded to somatosensory (median and radial nerve) stimulation. Four patients with cerebellar disease were tested and two showed abnormal VsCEPs with initial negativities. VsCEPs show distinct mapping over the posterior fossa and are likely to reflect climbing fibre responses via crossed otolith-cerebellar pathways.